![Fig 1. What can consumer wearables do? Heart rate can be measured with an oximeter built into a ring [3], muscle activity with an electromyographi
sensor embedded into clothing [4], stress with an electodermal sensor incorporated into a wristband [5], and physical activity or sleep patterns via an
accelerometer in a watch [6,7]. In addition, a female’s most fertile period can be identified with detailed body temperature tracking [8], while levels of me
attention can be monitored with a small number of non-gelled electroencephalogram (EEG) electrodes [9]. Levels of social interaction (also known to a
PLOS Medicine 2016](https://image.slidesharecdn.com/merck1hrv16digitalmedicine180115-180115023755/85/slide-65-320.jpg)
![DeepFace: Closing the Gap to Human-Level
Performance in FaceVerification
Taigman,Y. et al. (2014). DeepFace: Closing the Gap to Human-Level Performance in FaceVerification, CVPR’14.
Figure 2. Outline of the DeepFace architecture. A front-end of a single convolution-pooling-convolution filtering on the rectified input, followed by three
locally-connected layers and two fully-connected layers. Colors illustrate feature maps produced at each layer. The net includes more than 120 million
parameters, where more than 95% come from the local and fully connected layers.
very few parameters. These layers merely expand the input
into a set of simple local features.
The subsequent layers (L4, L5 and L6) are instead lo-
cally connected [13, 16], like a convolutional layer they ap-
ply a filter bank, but every location in the feature map learns
a different set of filters. Since different regions of an aligned
image have different local statistics, the spatial stationarity
The goal of training is to maximize the probability of
the correct class (face id). We achieve this by minimiz-
ing the cross-entropy loss for each training sample. If k
is the index of the true label for a given input, the loss is:
L = log pk. The loss is minimized over the parameters
by computing the gradient of L w.r.t. the parameters and
Human: 95% vs. DeepFace in Facebook: 97.35%
Recognition Accuracy for Labeled Faces in the Wild (LFW) dataset (13,233 images, 5,749 people)](https://image.slidesharecdn.com/merck1hrv16digitalmedicine180115-180115023755/85/slide-204-320.jpg)
![FaceNet:A Unified Embedding for Face
Recognition and Clustering
Schroff, F. et al. (2015). FaceNet:A Unified Embedding for Face Recognition and Clustering
Human: 95% vs. FaceNet of Google: 99.63%
Recognition Accuracy for Labeled Faces in the Wild (LFW) dataset (13,233 images, 5,749 people)
False accept
False reject
s. This shows all pairs of images that were
on LFW. Only eight of the 13 errors shown
he other four are mislabeled in LFW.
on Youtube Faces DB
ge similarity of all pairs of the first one
our face detector detects in each video.
False accept
False reject
Figure 6. LFW errors. This shows all pairs of images that were
incorrectly classified on LFW. Only eight of the 13 errors shown
here are actual errors the other four are mislabeled in LFW.
5.7. Performance on Youtube Faces DB
We use the average similarity of all pairs of the first one
hundred frames that our face detector detects in each video.
This gives us a classification accuracy of 95.12%±0.39.
Using the first one thousand frames results in 95.18%.
Compared to [17] 91.4% who also evaluate one hundred
frames per video we reduce the error rate by almost half.
DeepId2+ [15] achieved 93.2% and our method reduces this
error by 30%, comparable to our improvement on LFW.
5.8. Face Clustering
Our compact embedding lends itself to be used in order
to cluster a users personal photos into groups of people with
the same identity. The constraints in assignment imposed
by clustering faces, compared to the pure verification task,
lead to truly amazing results. Figure 7 shows one cluster in
a users personal photo collection, generated using agglom-
erative clustering. It is a clear showcase of the incredible
invariance to occlusion, lighting, pose and even age.
Figure 7. Face Clustering. Shown is an exemplar cluster for one
user. All these images in the users personal photo collection were
clustered together.
6. Summary
We provide a method to directly learn an embedding into
an Euclidean space for face verification. This sets it apart
from other methods [15, 17] who use the CNN bottleneck
layer, or require additional post-processing such as concate-
nation of multiple models and PCA, as well as SVM clas-
sification. Our end-to-end training both simplifies the setup
and shows that directly optimizing a loss relevant to the task
at hand improves performance.
Another strength of our model is that it only requires
False accept
False reject
Figure 6. LFW errors. This shows all pairs of images that were
incorrectly classified on LFW. Only eight of the 13 errors shown
here are actual errors the other four are mislabeled in LFW.
5.7. Performance on Youtube Faces DB
We use the average similarity of all pairs of the first one
hundred frames that our face detector detects in each video.
This gives us a classification accuracy of 95.12%±0.39.
Using the first one thousand frames results in 95.18%.
Compared to [17] 91.4% who also evaluate one hundred
frames per video we reduce the error rate by almost half.
DeepId2+ [15] achieved 93.2% and our method reduces this
error by 30%, comparable to our improvement on LFW.
5.8. Face Clustering
Our compact embedding lends itself to be used in order
to cluster a users personal photos into groups of people with
the same identity. The constraints in assignment imposed
by clustering faces, compared to the pure verification task,
Figure 7. Face Clustering. Shown is an exemplar cluster for one
user. All these images in the users personal photo collection were
clustered together.
6. Summary
We provide a method to directly learn an embedding into
an Euclidean space for face verification. This sets it apart
from other methods [15, 17] who use the CNN bottleneck
layer, or require additional post-processing such as concate-
nation of multiple models and PCA, as well as SVM clas-](https://image.slidesharecdn.com/merck1hrv16digitalmedicine180115-180115023755/85/slide-205-320.jpg)
![Targeting Ultimate Accuracy: Face
Recognition via Deep Embedding
Jingtuo Liu (2015) Targeting Ultimate Accuracy: Face Recognition via Deep Embedding
Human: 95% vs.Baidu: 99.77%
Recognition Accuracy for Labeled Faces in the Wild (LFW) dataset (13,233 images, 5,749 people)
3
Although several algorithms have achieved nearly perfect
accuracy in the 6000-pair verification task, a more practical
can achieve 95.8% identification rate, relatively reducing the
error rate by about 77%.
TABLE 3. COMPARISONS WITH OTHER METHODS ON SEVERAL EVALUATION TASKS
Score = -0.060 (pair #113) Score = -0.022 (pair #202) Score = -0.034 (pair #656)
Score = -0.031 (pair #1230) Score = -0.073 (pair #1862) Score = -0.091(pair #2499)
Score = -0.024 (pair #2551) Score = -0.036 (pair #2552) Score = -0.089 (pair #2610)
Method
Performance on tasks
Pair-wise
Accuracy(%)
Rank-1(%)
DIR(%) @
FAR =1%
Verification(%
)@ FAR=0.1%
Open-set
Identification(%
)@ Rank =
1,FAR = 0.1%
IDL Ensemble
Model
99.77 98.03 95.8 99.41 92.09
IDL Single Model 99.68 97.60 94.12 99.11 89.08
FaceNet[12] 99.63 NA NA NA NA
DeepID3[9] 99.53 96.00 81.40 NA NA
Face++[2] 99.50 NA NA NA NA
Facebook[15] 98.37 82.5 61.9 NA NA
Learning from
Scratch[4]
97.73 NA NA 80.26 28.90
HighDimLBP[10] 95.17 NA NA
41.66(reported
in [4])
18.07(reported
in [4])
• 6,000쌍의 얼굴 사진 중에 바이두의 인공지능은 불과 14쌍만을 잘못 판단
• 알고 보니 이 14쌍 중의 5쌍의 사진은 오히려 정답에 오류가 있었고,
실제로는 인공지능이 정확 (red box)](https://image.slidesharecdn.com/merck1hrv16digitalmedicine180115-180115023755/85/slide-206-320.jpg)
![AtomNet: A Deep Convolutional Neural Network for
Bioactivity Prediction in Structure-based Drug
Discovery
Izhar Wallach
Atomwise, Inc.
izhar@atomwise.com
Michael Dzamba
Atomwise, Inc.
misko@atomwise.com
Abraham Heifets
Atomwise, Inc.
abe@atomwise.com
Abstract
Deep convolutional neural networks comprise a subclass of deep neural networks
(DNN) with a constrained architecture that leverages the spatial and temporal
structure of the domain they model. Convolutional networks achieve the best pre-
dictive performance in areas such as speech and image recognition by hierarchi-
cally composing simple local features into complex models. Although DNNs have
been used in drug discovery for QSAR and ligand-based bioactivity predictions,
none of these models have benefited from this powerful convolutional architec-
ture. This paper introduces AtomNet, the first structure-based, deep convolutional
neural network designed to predict the bioactivity of small molecules for drug dis-
covery applications. We demonstrate how to apply the convolutional concepts of
feature locality and hierarchical composition to the modeling of bioactivity and
chemical interactions. In further contrast to existing DNN techniques, we show
that AtomNet’s application of local convolutional filters to structural target infor-
mation successfully predicts new active molecules for targets with no previously
known modulators. Finally, we show that AtomNet outperforms previous docking
approaches on a diverse set of benchmarks by a large margin, achieving an AUC
greater than 0.9 on 57.8% of the targets in the DUDE benchmark.
1 Introduction
Fundamentally, biological systems operate through the physical interaction of molecules. The ability
to determine when molecular binding occurs is therefore critical for the discovery of new medicines
and for furthering of our understanding of biology. Unfortunately, despite thirty years of compu-
tational efforts, computer tools remain too inaccurate for routine binding prediction, and physical
experiments remain the state of the art for binding determination. The ability to accurately pre-
dict molecular binding would reduce the time-to-discovery of new treatments, help eliminate toxic
molecules early in development, and guide medicinal chemistry efforts [1, 2].
In this paper, we introduce a new predictive architecture, AtomNet, to help address these challenges.
AtomNet is novel in two regards: AtomNet is the first deep convolutional neural network for molec-
ular binding affinity prediction. It is also the first deep learning system that incorporates structural
information about the target to make its predictions.
Deep convolutional neural networks (DCNN) are currently the best performing predictive models
for speech and vision [3, 4, 5, 6]. DCNN is a class of deep neural network that constrains its model
architecture to leverage the spatial and temporal structure of its domain. For example, a low-level
image feature, such as an edge, can be described within a small spatially-proximate patch of pixels.
Such a feature detector can share evidence across the entire receptive field by “tying the weights”
of the detector neurons, as the recognition of the edge does not depend on where it is found within
1
arXiv:1510.02855v1[cs.LG]10Oct2015](https://image.slidesharecdn.com/merck1hrv16digitalmedicine180115-180115023755/85/slide-223-320.jpg)
![AtomNet: A Deep Convolutional Neural Network for
Bioactivity Prediction in Structure-based Drug
Discovery
Izhar Wallach
Atomwise, Inc.
izhar@atomwise.com
Michael Dzamba
Atomwise, Inc.
misko@atomwise.com
Abraham Heifets
Atomwise, Inc.
abe@atomwise.com
Abstract
Deep convolutional neural networks comprise a subclass of deep neural networks
(DNN) with a constrained architecture that leverages the spatial and temporal
structure of the domain they model. Convolutional networks achieve the best pre-
dictive performance in areas such as speech and image recognition by hierarchi-
cally composing simple local features into complex models. Although DNNs have
been used in drug discovery for QSAR and ligand-based bioactivity predictions,
none of these models have benefited from this powerful convolutional architec-
ture. This paper introduces AtomNet, the first structure-based, deep convolutional
neural network designed to predict the bioactivity of small molecules for drug dis-
covery applications. We demonstrate how to apply the convolutional concepts of
feature locality and hierarchical composition to the modeling of bioactivity and
chemical interactions. In further contrast to existing DNN techniques, we show
that AtomNet’s application of local convolutional filters to structural target infor-
mation successfully predicts new active molecules for targets with no previously
known modulators. Finally, we show that AtomNet outperforms previous docking
approaches on a diverse set of benchmarks by a large margin, achieving an AUC
greater than 0.9 on 57.8% of the targets in the DUDE benchmark.
1 Introduction
Fundamentally, biological systems operate through the physical interaction of molecules. The ability
to determine when molecular binding occurs is therefore critical for the discovery of new medicines
and for furthering of our understanding of biology. Unfortunately, despite thirty years of compu-
tational efforts, computer tools remain too inaccurate for routine binding prediction, and physical
experiments remain the state of the art for binding determination. The ability to accurately pre-
dict molecular binding would reduce the time-to-discovery of new treatments, help eliminate toxic
molecules early in development, and guide medicinal chemistry efforts [1, 2].
In this paper, we introduce a new predictive architecture, AtomNet, to help address these challenges.
AtomNet is novel in two regards: AtomNet is the first deep convolutional neural network for molec-
ular binding affinity prediction. It is also the first deep learning system that incorporates structural
information about the target to make its predictions.
Deep convolutional neural networks (DCNN) are currently the best performing predictive models
for speech and vision [3, 4, 5, 6]. DCNN is a class of deep neural network that constrains its model
architecture to leverage the spatial and temporal structure of its domain. For example, a low-level
image feature, such as an edge, can be described within a small spatially-proximate patch of pixels.
Such a feature detector can share evidence across the entire receptive field by “tying the weights”
of the detector neurons, as the recognition of the edge does not depend on where it is found within
1
arXiv:1510.02855v1[cs.LG]10Oct2015
Smina 123 35 5 0 0
Table 3: The number of targets on which AtomNet and Smina exceed given adjusted-logAUC thresh-
olds. For example, on the CHEMBL-20 PMD set, AtomNet achieves an adjusted-logAUC of 0.3
or better for 27 targets (out of 50 possible targets). ChEMBL-20 PMD contains 50 targets, DUDE-
30 contains 30 targets, DUDE-102 contains 102 targets, and ChEMBL-20 inactives contains 149
targets.
To overcome these limitations we take an indirect approach. Instead of directly visualizing filters
in order to understand their specialization, we apply filters to input data and examine the location
where they maximally fire. Using this technique we were able to map filters to chemical functions.
For example, Figure 5 illustrate the 3D locations at which a particular filter from our first convo-
lutional layer fires. Visual inspection of the locations at which that filter is active reveals that this
filter specializes as a sulfonyl/sulfonamide detector. This demonstrates the ability of the model to
learn complex chemical features from simpler ones. In this case, the filter has inferred a meaningful
spatial arrangement of input atom types without any chemical prior knowledge.
Figure 5: Sulfonyl/sulfonamide detection with autonomously trained convolutional filters.
8
• 이미 알려진 단백질-리간드 3차원 결합 구조를 딥러닝(CNN)으로 학습
• 화학 결합 등에 대한 계산 없이도, 단백질-리간드 결합 여부를 계산
• 기존의 구조기반 예측 등 대비, 딥러닝으로 더 정확히 예측하였음](https://image.slidesharecdn.com/merck1hrv16digitalmedicine180115-180115023755/85/slide-224-320.jpg)
![AtomNet: A Deep Convolutional Neural Network for
Bioactivity Prediction in Structure-based Drug
Discovery
Izhar Wallach
Atomwise, Inc.
izhar@atomwise.com
Michael Dzamba
Atomwise, Inc.
misko@atomwise.com
Abraham Heifets
Atomwise, Inc.
abe@atomwise.com
Abstract
Deep convolutional neural networks comprise a subclass of deep neural networks
(DNN) with a constrained architecture that leverages the spatial and temporal
structure of the domain they model. Convolutional networks achieve the best pre-
dictive performance in areas such as speech and image recognition by hierarchi-
cally composing simple local features into complex models. Although DNNs have
been used in drug discovery for QSAR and ligand-based bioactivity predictions,
none of these models have benefited from this powerful convolutional architec-
ture. This paper introduces AtomNet, the first structure-based, deep convolutional
neural network designed to predict the bioactivity of small molecules for drug dis-
covery applications. We demonstrate how to apply the convolutional concepts of
feature locality and hierarchical composition to the modeling of bioactivity and
chemical interactions. In further contrast to existing DNN techniques, we show
that AtomNet’s application of local convolutional filters to structural target infor-
mation successfully predicts new active molecules for targets with no previously
known modulators. Finally, we show that AtomNet outperforms previous docking
approaches on a diverse set of benchmarks by a large margin, achieving an AUC
greater than 0.9 on 57.8% of the targets in the DUDE benchmark.
1 Introduction
Fundamentally, biological systems operate through the physical interaction of molecules. The ability
to determine when molecular binding occurs is therefore critical for the discovery of new medicines
and for furthering of our understanding of biology. Unfortunately, despite thirty years of compu-
tational efforts, computer tools remain too inaccurate for routine binding prediction, and physical
experiments remain the state of the art for binding determination. The ability to accurately pre-
dict molecular binding would reduce the time-to-discovery of new treatments, help eliminate toxic
molecules early in development, and guide medicinal chemistry efforts [1, 2].
In this paper, we introduce a new predictive architecture, AtomNet, to help address these challenges.
AtomNet is novel in two regards: AtomNet is the first deep convolutional neural network for molec-
ular binding affinity prediction. It is also the first deep learning system that incorporates structural
information about the target to make its predictions.
Deep convolutional neural networks (DCNN) are currently the best performing predictive models
for speech and vision [3, 4, 5, 6]. DCNN is a class of deep neural network that constrains its model
architecture to leverage the spatial and temporal structure of its domain. For example, a low-level
image feature, such as an edge, can be described within a small spatially-proximate patch of pixels.
Such a feature detector can share evidence across the entire receptive field by “tying the weights”
of the detector neurons, as the recognition of the edge does not depend on where it is found within
1
arXiv:1510.02855v1[cs.LG]10Oct2015
• 이미 알려진 단백질-리간드 3차원 결합 구조를 딥러닝(CNN)으로 학습
• 화학 결합 등에 대한 계산 없이도, 단백질-리간드 결합 여부를 계산
• 기존의 구조기반 예측 등 대비, 딥러닝으로 더 정확히 예측하였음](https://image.slidesharecdn.com/merck1hrv16digitalmedicine180115-180115023755/85/slide-225-320.jpg)
Uploaded byYoon Sup Choi
의료의 미래, 디지털 헬스케어: 제약산업을 중심으로
Professor Yoon Sup Choi discusses digital health and the future of healthcare centered around changes in the pharmaceutical industry. He notes that three key steps in implementing digital medicine are: 1) measuring data through devices like smartphones, wearables, and genetic analysis; 2) collecting the data; and 3) gaining insights from the data using artificial intelligence. Choi also provides an overview of the digital health industry landscape and increasing investment in digital health startups from pharmaceutical companies and other investors.
![[365mc] 디지털 헬스케어: 의료의 미래](https://cdn.slidesharecdn.com/ss_thumbnails/365mc-200317140059-thumbnail.jpg?width=640&height=640&fit=bounds)
![[C&C] 의료의 미래 디지털 헬스케어](https://cdn.slidesharecdn.com/ss_thumbnails/cc-200616080105-thumbnail.jpg?width=640&height=640&fit=bounds)
![[KNAPS] 포스트 코로나 시대, 제약 산업과 디지털 헬스케어](https://cdn.slidesharecdn.com/ss_thumbnails/knaps-210128022335-thumbnail.jpg?width=640&height=640&fit=bounds)
![[NAVER D2SF][TMS2019] 휴이노](https://cdn.slidesharecdn.com/ss_thumbnails/trackb03huinnofinweb-191113095606-thumbnail.jpg?width=640&height=640&fit=bounds)
![[대한병리학회] 의료 인공지능 101: 병리를 중심으로](https://cdn.slidesharecdn.com/ss_thumbnails/pathology-201106004112-thumbnail.jpg?width=640&height=640&fit=bounds)
![[ASGO 2019] Artificial Intelligence in Medicine](https://cdn.slidesharecdn.com/ss_thumbnails/asgo2019medicalai191010-191010155811-thumbnail.jpg?width=640&height=640&fit=bounds)
의료의 미래, 디지털 헬스케어: 제약산업을 중심으로
- 1. Professor, SAHIST, SungkyunkwanUniversity Director, Digital Healthcare Institute Yoon Sup Choi, Ph.D. 의료의 미래, 디지털 헬스케어 제약 산업의 변화를 중심으로
- 2. “It's in Apple'sDNA that technology alone is not enough. It's technology married with liberal arts.”
- 3. The Convergence ofIT, BT and Medicine
- 6. Inevitable Tsunami ofChange
- 7.
- 8. •2017년은 역대 디지털헬스케어 스타트업 펀딩 중 최대의 해. •투자횟수와 개별 투자의 규모도 역대 최고 수준을 기록 •$100m 을 넘는 mega deal 도 8건이 있었으며, •이에 따라 기업가치 $1b이 넘는 유니콘 기업들이 상당수 생겨남. https://rockhealth.com/reports/2017-year-end-funding-report-the-end-of-the-beginning-of-digital-health/
- 9. 2010 2011 20122013 2014 2015 2016 2017 Q1 Q2 Q3 Q4 FUNDING SNAPSHOT: YEAR OVER YEAR 6 155 284 477 647 596 550 658 794 Deal Count $1.4B $1.7B $1.7B $627M $603M$459M $288M $8.2B $6.2B $7.2B $2.9B $2.3B $2.0B $1.2B $11.5B $2.3B 2017 was the most active year for digital health funding to date with more than $11.5B invested across a record-setting 794 deals. Q4 2017 also had record-breaking numbers, surpassing $2B across 227 deals (the most ever in one quarter). Given the global market opportunity, increasing demand for innovation, wave of high-quality entrepreneurs flocking to the sector, and early stage of this innovation cycle, we expect plentiful capital in 2018. DEALS & FUNDING OUTLOOKGEOGRAPHY INVESTORS Source: StartUp Health Insights | startuphealth.com/insights Note: Report based on public data through 12/31/17 on seed (incl. accelerator), venture, corporate venture and private equity funding only. © 2018 StartUp Health LLC https://www.slideshare.net/StartUpHealth/2017-startup-health-insights-year-end-report
- 10. https://rockhealth.com/reports/digital-health-funding-2015-year-in-review/ •Merck는 CVC 중에디지털 헬스케어 분야 투자가 가장 활발한 펀드 중 하나 •2011-2015년 20건의 투자 집행으로 CVC 중 1위
- 11. THE MOST ACTIVEINVESTORS OF 2017 21 *Deal counts excludes accelerator rounds DEALS & FUNDING GEOGRAPHY INVESTORS OUTLOOK Khosla Ventures and GE Ventures were once again part of the top three most active investors as they were in 2016, though their deal count decreased. Familiar VCs like Sequoia, NEA, Norwest, and Safeguard all increased their deal activity substantially year over year. Source: StartUp Health Insights | startuphealth.com/insights Note: Report based on public data through 12/31/17 on seed (incl. accelerator), venture, corporate venture and private equity funding only. © 2018 StartUp Health LLC Firm 2017 Deals* 2016 Deals* Stage Early Mid Late 1 10 15 2 9 5 2 9 18 2 9 5 2 9 5 6 8 3 6 8 4 6 8 4 9 7 7 9 7 5 9 7 2 9 7 4 https://www.slideshare.net/StartUpHealth/2017-startup-health-insights-year-end-report •Merck는 2017년 글로벌 디지털 헬스케어 스타트업 투자에서도 8건으로 선두권 (1위 10건)
- 12. •최근 3년 동안Merck, J&J, GSK 등의 제약사들의 디지털 헬스케어 분야 투자 급증 •2015-2016년 총 22건의 deal (=2010-2014년의 5년간 투자 건수와 동일) •Merck 가 가장 활발: 2009년부터 Global Health Innovation Fund 를 통해 24건 투자 ($5-7M) •GSK 의 경우 2014년부터 6건 (via VC arm, SR One): including Propeller Health
- 14. 헬스케어넓은 의미의 건강관리에는 해당되지만, 디지털 기술이 적용되지 않고, 전문 의료 영역도 아닌 것 예) 운동, 영양, 수면 디지털 헬스케어 건강 관리 중에 디지털 기술이 사용되는 것 예) 사물인터넷, 인공지능, 3D 프린터, VR/AR 모바일 헬스케어 디지털 헬스케어 중 모바일 기술이 사용되는 것 예) 스마트폰, 사물인터넷, SNS 개인 유전정보분석 예) 암유전체, 질병위험도, 보인자, 약물 민감도 예) 웰니스, 조상 분석 헬스케어 관련 분야 구성도(ver 0.3) 의료 질병 예방, 치료, 처방, 관리 등 전문 의료 영역 원격의료 원격진료
- 15. What is mostimportant factor in digital medicine?
- 16. “Data! Data! Data!”he cried.“I can’t make bricks without clay!” - Sherlock Holmes,“The Adventure of the Copper Beeches”
- 18. 새로운 데이터가 새로운 방식으로 새로운주체에 의해 측정, 저장, 통합, 분석된다. 데이터의 종류 데이터의 질적/양적 측면 웨어러블 기기 스마트폰 유전 정보 분석 인공지능 SNS 사용자/환자 대중
- 19. Three Steps toImplement Digital Medicine • Step 1. Measure the Data • Step 2. Collect the Data • Step 3. Insight from the Data
- 20. Digital Healthcare IndustryLandscape Data Measurement Data Integration Data Interpretation Treatment Smartphone Gadget/Apps DNA Artificial Intelligence 2nd Opinion Wearables / IoT (ver. 3) EMR/EHR 3D Printer Counseling Data Platform Accelerator/early-VC Telemedicine Device On Demand (O2O) VR Digital Healthcare Institute Diretor, Yoon Sup Choi, Ph.D. yoonsup.choi@gmail.com
- 21. Data Measurement DataIntegration Data Interpretation Treatment Smartphone Gadget/Apps DNA Artificial Intelligence 2nd Opinion Device On Demand (O2O) Wearables / IoT Digital Healthcare Institute Diretor, Yoon Sup Choi, Ph.D. yoonsup.choi@gmail.com EMR/EHR 3D Printer Counseling Data Platform Accelerator/early-VC VR Telemedicine Digital Healthcare Industry Landscape (ver. 3)
- 22. Step 1. Measurethe Data
- 23. Smartphone: the originof healthcare innovation
- 24. Smartphone: the originof healthcare innovation
- 25. 2013? The election ofPope Benedict The Election of Pope Francis
- 26. The Election ofPope Francis The Election of Pope Benedict
- 27.
- 30. Sensor Growth inSmartphones https://www.qualcomm.com/news/onq/2014/04/24/behind-sixth-sense-smartphones-snapdragon-processor-sensor-engine
- 31. Sci Transl Med2015
- 32.
- 33.
- 34.
- 35.
- 37. Smartphone video microscope automatesdetection of parasites in blood
- 38.
- 39. AliveCor Heart Monitor(Kardia)
- 40. AliveCor Heart Monitor(Kardia)
- 43.
- 45. BeyondVerbal: Reading emotionsfrom voices
- 46.
- 47. BeyondVerbal • 기계가 사람의감정을 이해한다면? • 헬스케어 분야에서도 응용도 높음: 슬픔/우울함/피로 등의 감정 파악 • 일부 보험 회사에서는 가입자의 우울증 여부 파악을 위해 이미 사용 중 • Aetna 는 2012년 부터 고객의 우울증 여부를 전화 목소리 분석으로 파악 • 기존의 방식에 비해 우울증 환자 6배 파악 • 사생활 침해 여부 존재
- 49. Digital Phenotype: Your smartphoneknows if you are depressed Ginger.io
- 50. Digital Phenotype: Your smartphoneknows if you are depressed J Med Internet Res. 2015 Jul 15;17(7):e175. The correlation analysis between the features and the PHQ-9 scores revealed that 6 of the 10 features were significantly correlated to the scores: • strong correlation: circadian movement, normalized entropy, location variance • correlation: phone usage features, usage duration and usage frequency
- 52. • 아이폰의 센서로측정한 자신의 의료/건강 데이터를 플랫폼에 공유 가능 • 가속도계, 마이크, 자이로스코프, GPS 센서 등을 이용 • 걸음, 운동량, 기억력, 목소리 떨림 등등 • 기존의 의학연구의 문제를 해결: 충분한 의료 데이터의 확보 • 연구 참여자 등록에 물리적, 시간적 장벽을 제거 (1번/3개월 ➞ 1번/1초) • 대중의 의료 연구 참여 장려: 연구 참여자의 수 증가 • 발표 후 24시간 내에 수만명의 연구 참여자들이 지원 • 사용자 본인의 동의 하에 진행 Research Kit
- 53. •초기 버전으로, 5가지질환에 대한 앱 5개를 소개 ResearchKit
- 54.
- 55.
- 56.
- 57. http://www.roche.com/media/store/roche_stories/roche-stories-2015-08-10.htm pRED app totrack Parkinson’s symptoms in drug trial
- 58. Autism and BeyondEpiWatchMole Mapper measuring facial expressions of young patients having autism measuring morphological changes of moles measuring behavioral data of epilepsy patients
- 59. •스탠퍼드의 심혈관 질환연구 앱, myHeart • 발표 하루만에 11,000 명의 참가자가 등록 • 스탠퍼드의 해당 연구 책임자 앨런 영, “기존의 방식으로는 11,000명 참가자는 미국 전역의 50개 병원에서 1년간 모집해야 한다”
- 60. •파킨슨 병 연구앱, mPower • 발표 하루만에 5,589 명의 참가자가 등록 • 기존에 6000만불을 들여 5년 동안 모집한 환자의 수는 단 800명
- 61.
- 64. http://www.rolls-royce.com/about/our-technology/enabling-technologies/engine-health-management.aspx#sense 250 sensors tomonitor the “health” of the GE turbines
- 65. Fig 1. Whatcan consumer wearables do? Heart rate can be measured with an oximeter built into a ring [3], muscle activity with an electromyographi sensor embedded into clothing [4], stress with an electodermal sensor incorporated into a wristband [5], and physical activity or sleep patterns via an accelerometer in a watch [6,7]. In addition, a female’s most fertile period can be identified with detailed body temperature tracking [8], while levels of me attention can be monitored with a small number of non-gelled electroencephalogram (EEG) electrodes [9]. Levels of social interaction (also known to a PLOS Medicine 2016
- 66. PwC Health ResearchInstitute Health wearables: Early days2 insurers—offering incentives for use may gain traction. HRI’s survey Source: HRI/CIS Wearables consumer survey 2014 21% of US consumers currently own a wearable technology product 2% wear it a few times a month 2% no longer use it 7% wear it a few times a week 10% wear it everyday Figure 2: Wearables are not mainstream – yet Just one in five US consumers say they own a wearable device. Intelligence Series sought to better understand American consumers’ attitudes toward wearables through done with the data. PwC, Health wearables: early days, 2014
- 67. PwC | TheWearable Life | 3 device (up from 21% in 2014). And 36% own more than one. We didn’t even ask this question in our previous survey since it wasn’t relevant at the time. That’s how far we’ve come. millennials are far more likely to own wearables than older adults. Adoption of wearables declines with age. Of note in our survey findings, however: Consumers aged 35 to 49 are more likely to own smart watches. Across the board for gender, age, and ethnicity, fitness wearable technology is most popular. Fitness band Smart clothing Smart video/ photo device (e.g. GoPro) Smart watch Smart glasses* 45% 14% 27% 15% 12% Base: Respondents who currently own at least one device (pre-quota sample, n=700); Q10A/B/C/D/E. Please tell us your relationship with the following wearable technology products. *Includes VR/AR glasses Fitness runs away with it % respondents who own type of wearable device PwC,The Wearable Life 2.0, 2016 • 49% own at least one wearable device (up from 21% in2014) • 36% own more than one device.
- 68.
- 69.
- 70. https://clinicaltrials.gov/ct2/results?term=fitbit&Search=Search •의료기기가 아님에도 Fitbit은 이미 임상 연구에 폭넓게 사용되고 있음 •Fitbit 이 장려하지 않았음에도, 임상 연구자들이 자발적으로 사용 •Fitbit 을 이용한 임상 연구 수는 계속 증가하는 추세 (16.3(80), 16.8(113), 17.7(173))
- 72. •Fitbit이 임상연구에 활용되는것은 크게 두 가지 경우 •Fitbit 자체가 intervention이 되어서 활동량이나 치료 효과를 증진시킬 수 있는지 여부 •연구 참여자의 활동량을 모니터링 하기 위한 수단 •1. Fitbit으로 환자의 활동량을 증가시키기 위한 연구들 •Fitbit이 소아 비만 환자의 활동량을 증가시키는지 여부를 연구 •Fitbit이 위소매절제술을 받은 환자들의 활동량을 증가시키는지 여부 •Fitbit이 젊은 낭성 섬유증 (cystic fibrosis) 환자의 활동량을 증가시키는지 여부 •Fitbit이 암 환자의 신체 활동량을 증가시키기 위한 동기부여가 되는지 여부 •2. Fitbit으로 임상 연구에 참여하는 환자의 활동량을 모니터링 •항암 치료를 받은 환자들의 건강과 예후를 평가하는데 fitbit을 사용 •현금이 자녀/부모의 활동량을 증가시키는지 파악하기 위해 fitbit을 사용 •Brain tumor 환자의 삶의 질 측정을 위해 다른 survey 결과와 함께 fitbit을 사용 •말초동맥 질환(Peripheral Artery Disease) 환자의 활동량을 평가하기 위해
- 73. •체중 감량이 유방암재발에 미치는 영향을 연구 •유방암 환자들 중 20%는 재발, 대부분이 전이성 유방암 •과체중은 유방암의 위험을 높인다고 알려져 왔으며, •비만은 초기 유방암 환자의 예후를 좋지 않게 만드는 것도 알려짐 •하지만, 체중 감량과 유방암 재발 위험도의 상관관계 연구는 아직 없음 •3,200 명의 과체중, 초기 비만 유방암 환자들이 2년간 참여 •결과에 따라 전세계 유방암 환자의 표준 치료에 체중 감량이 포함될 가능성 •Fitbit 이 체중 감량 프로그램에 대한 지원 •Fitbit Charge HR: 운동량, 칼로리 소모, 심박수 측정 •Fitbit Aria Wi-Fi Smart Scale: 스마트 체중계 •FitStar: 개인 맞춤형 동영상 운동 코칭 서비스 2016. 4. 27.
- 75.
- 76. •Biogen Idec, 다발성경화증 환자의 모니터링에 Fitbit을 사용 •고가의 약 효과성을 검증하여 보험 약가 유지 목적 •정교한 측정으로 MS 전조 증상의 조기 발견 가능? Dec 23, 2014
- 77.
- 82. (“FREE VERTICAL MOMENTSAND TRANSVERSE FORCES IN HUMAN WALKING AND THEIR ROLE IN RELATION TO ARM-SWING”, YU LI*, WEIJIE WANG, ROBIN H. CROMPTON AND MICHAEL M. GUNTHER) (“SYNTHESIS OF NATURAL ARM SWING MOTION IN HUMAN BIPEDAL WALKING”, JAEHEUNG PARK) ︎ Right Arm Left Foot Left Arm Right Foot “보행 시 팔의 움직임은 몸의 역학적 균형을 맞추기 위한 자동적인 행동 으로, 반대쪽 발의 움직임을 관찰할 수 있는 지표” 보행 종류에 따른 신체 운동 궤도의 변화 발의 모양 팔의 스윙 궤도 일반 보행 팔자 걸음 구부린 걸음 직토 워크에서 수집하는 데이터 종류 설명 비고 충격량 발에 전해지는 충격량 분석 Impact Score 보행 주기 보행의 주기 분석 Interval Score 보폭 단위 보행 시의 거리 Stride(향후 보행 분석 고도화용) 팔의 3차원 궤도 걸음에 따른 팔의 움직임 궤도 팔의 Accel,Gyro Data 취합 보행 자세 상기 자료를 분석한 보행 자세 분류 총 8가지 종류로 구분 비대칭 지수 신체 부위별(어깨, 허리, 골반) 비대칭 점수 제공 1주일 1회 반대쪽 손 착용을 통한 데이터 취득 필요 걸음걸이 템플릿 보행시 발생하는 특이점들을 추출하여 개인별 템플릿 저장 생체 인증 기능용 with the courtesy of ZIKTO, Inc
- 83.
- 85.
- 86.
- 87.
- 88.
- 90. • $20 • thefirst and only 24-hour thermometer • constantly monitor baby’s temperature • FDA cleared
- 91.
- 92.
- 96.
- 99. Withings Wireless BloodPressure Monitor
- 100. Huinno: Cuff-less BloodPressure Monitor
- 101. Ingestible Sensor, ProteusDigital Health
- 102. Ingestible Sensor, ProteusDigital Health
- 104. IEEE Trans BiomedEng. 2014 Jul An Ingestible Sensor for Measuring Medication Adherence d again on imal was ysis were s detected, risk of ed with a his can be s during can be on, placed filling, or an edible monstrated cases, the nts of the ve release ity, visual a suitable The 0.9% of devices that went undetected represent contributions from all components of the system. For the sensor, the most likely contribution is due to physiological corner cases, where a combination of stomach environment and receiver-sensor orientation may result in a small proportion of devices (no greater than 0.9%) being missed. Table IV- Exposure and performance in clinical trials 412 subjects 20,993 ingestions Maximum daily ingestion: 34 Maximum use days: 90 days 99.1% Detection accuracy 100% Correct identification 0% False positives No SAEs / UADEs related to system Trials were conducted in the following patient populations. The number of patients in each study is indicated in parentheses: Healthy Volunteers (296), Cardiovascular disease (53), Tuberculosis (30), Psychiatry (28). SAE = Serious Adverse Event; UADE = Unanticipated Adverse Device Effect) Exposure and performance in clinical trials
- 105. Jan 12, 2015 Clinicaltrial researchers using Oracle’s software will now be able to track patients’ medication adherence with Proteus’s technology. - Measuring participant adherence to drug protocols - Identifying the optimum dosing regimen for recommended use
- 106. Sep 10, 2015 Proteusand Otsuka have submitted a sensor-embedded version of the antidepressant Abilify for FDA approval.
- 107.
- 108. Nov 13, 2017 •2017년11월 FDA는 Abilify MyCite의 시판 허가 •처방 전 환자의 동의가 필요 •환자의 사생활 침해 우려 의견도 있음 •주치의와 보호자까지 최대 4명이 복약 정보 수령 가능
- 109. Nov 13, 2017 •2017년11월 FDA는 Abilify MyCite의 시판 허가 •처방 전 환자의 동의가 필요 •환자의 사생활 침해 우려 의견도 있음 •주치의와 보호자까지 최대 4명이 복약 정보 수령 가능
- 110.
- 111.
- 112. 2003 Human GenomeProject 13 years (676 weeks) $2,700,000,000 2007 Dr. CraigVenter’s genome 4 years (208 weeks) $100,000,000 2008 Dr. James Watson’s genome 4 months (16 weeks) $1,000,000 2009 (Nature Biotechnology) 4 weeks $48,000 2013 1-2 weeks ~$5,000
- 113. 13 years 30hours (676 weeks) Over the last decade,
- 114. $2,700,000,000 ~$1,000 Over thelast decade,
- 115. Ferrari 458 Spider $398,00040 cents http://www.guardian.co.uk/science/2013/jun/08/genome-sequenced
- 116. The $1000 Genomeis Already Here!
- 117. •2017년 1월 NovaSeq5000, 6000 발표 •몇년 내로 $100로 WES 를 실현하겠다고 공언 •2일에 60명의 WES 가능 (한 명당 한 시간 이하)
- 122. Results within 6-8weeksA little spit is all it takes! DTC Genetic TestingDirect-To-Consumer
- 123. 120 Disease Risk 21Drug Response 49 Carrier Status 57Traits $99
- 124.
- 125.
- 126.
- 127.
- 128. Inherited Conditions 혈색소증은 유전적원인으로 철에 대한 체내 대사에 이상이 생겨 음식을 통해 섭취한 철이 너무 많이 흡수되는 질환입니다. 너무 많이 흡수된 철 은 우리 몸의 여러 장기, 특히 간, 심장 및 췌장에 과다하게 축적되며 이 들 장기를 손상시킴으로써 간질환, 심장질환 및 악성종양을 유발합니다.
- 129. Traits 음주 후 얼굴이붉어지는가 쓴 맛을 감지할 수 있나 귀지 유형 눈 색깔 곱슬머리 여부 유당 분해 능력 말라리아 저항성 대머리가 될 가능성 근육 퍼포먼스 혈액형 노로바이러스 저항성 HIV 저항성 흡연 중독 가능성
- 130.
- 131.
- 132. genetic factor vs.environmental factor
- 133. 1,200,000 1,000,000 900,000 850,000 650,000 500,000 400,000 300,000 250,000 180,000 100,000 2007-11 2011-06 2011-10 2012-04 2012-10 2013-04 2013-06 2013-09 2013-12 2014-10 2015-02 2015-05 2015-06 2016-02 0 Customer growth of23andMe 2017-04 2,000,000 Digital Healthcare Institute Director,Yoon Sup Choi, PhD yoonsup.choi@gmail.com
- 134. https://www.23andme.com/slideshow/research/ 고객의 자발적인 참여에의한 유전학 연구 깍지를 끼면 어느 쪽 엄지가 위로 오는가? 아침형 인간? 저녁형 인간? 빛에 노출되었을 때 재채기를 하는가? 근육의 퍼포먼스 쓴 맛 인식 능력 음주 후 얼굴이 붉어지나? 유당 분해 효소 결핍? 고객의 81%가 10개 이상의 질문에 자발적 답변 매주 1 million 개의 data point 축적 The More Data, The Higher Accuracy!
- 135. January 13, 2015January6, 2015 Data Business
- 136. Human genomes arebeing sequenced at an ever-increasing rate. The 1000 Genomes Project has aggregated hundreds of genomes; The Cancer Genome Atlas (TGCA) has gathered several thousand; and the Exome Aggregation Consortium (ExAC) has sequenced more than 60,000 exomes. Dotted lines show three possible future growth curves. DNA SEQUENCING SOARS 2001 2005 2010 2015 2020 2025 100 103 106 109 Human Genome Project Cumulativenumberofhumangenomes 1000 Genomes TCGA ExAC Current amount 1st personal genome Recorded growth Projection Double every 7 months (historical growth rate) Double every 12 months (Illumina estimate) Double every 18 months (Moore's law) Michael Einsetein, Nature, 2015
- 137. more rapid andaccurate approaches to infectious diseases. The driver mutations and key biologic unde Sequencing Applications in Medicine from Prewomb to Tomb Cell. 2014 Mar 27; 157(1): 241–253.
- 139. the manifestations ofdisease by providing a more comprehensive and nuanced view of the experience of illness. Through the lens of the digital phenotype, an individual’s interaction The digital phenotype Sachin H Jain, Brian W Powers, Jared B Hawkins & John S Brownstein In the coming years, patient phenotypes captured to enhance health and wellness will extend to human interactions with digital technology. In 1982, the evolutionary biologist Richard Dawkins introduced the concept of the “extended phenotype”1, the idea that pheno- types should not be limited just to biological processes, such as protein biosynthesis or tissue growth, but extended to include all effects that a gene has on its environment inside or outside ofthebodyoftheindividualorganism.Dawkins stressed that many delineations of phenotypes are arbitrary. Animals and humans can modify their environments, and these modifications andassociatedbehaviorsareexpressionsofone’s genome and, thus, part of their extended phe- notype. In the animal kingdom, he cites damn buildingbybeaversasanexampleofthebeaver’s extended phenotype1. Aspersonaltechnologybecomesincreasingly embedded in human lives, we think there is an important extension of Dawkins’s theory—the notion of a ‘digital phenotype’. Can aspects of ourinterfacewithtechnologybesomehowdiag- nosticand/orprognosticforcertainconditions? Can one’s clinical data be linked and analyzed together with online activity and behavior data to create a unified, nuanced view of human dis- ease?Here,wedescribetheconceptofthedigital phenotype. Although several disparate studies have touched on this notion, the framework for medicine has yet to be described. We attempt to define digital phenotype and further describe the opportunities and challenges in incorporat- ing these data into healthcare. Jan. 2013 0.000 0.002 0.004 Density 0.006 July 2013 Jan. 2014 July 2014 User 1 User 2 User 3 User 4 User 5 User 6 User 7 Date Figure 1 Timeline of insomnia-related tweets from representative individuals. Density distributions (probability density functions) are shown for seven individual users over a two-year period. Density on the y axis highlights periods of relative activity for each user. A representative tweet from each user is shown as an example. npg©2015NatureAmerica,Inc.Allrightsreserved. http://www.nature.com/nbt/journal/v33/n5/full/nbt.3223.html
- 140. ers, Jared BHawkins & John S Brownstein phenotypes captured to enhance health and wellness will extend to human interactions with st Richard pt of the hat pheno- biological sis or tissue effects that or outside m.Dawkins phenotypes can modify difications onsofone’s ended phe- cites damn thebeaver’s ncreasingly there is an heory—the aspects of ehowdiag- Jan. 2013 0.000 0.002 0.004 Density 0.006 July 2013 Jan. 2014 July 2014 User 1 User 2 User 3 User 4 User 5 User 6 User 7 Date Figure 1 Timeline of insomnia-related tweets from representative individuals. Density distributions (probability density functions) are shown for seven individual users over a two-year period. Density on the y axis highlights periods of relative activity for each user. A representative tweet from each user is Your twitter knows if you cannot sleep Timeline of insomnia-related tweets from representative individuals. Nat. Biotech. 2015
- 141. Reece & Danforth,“Instagram photos reveal predictive markers of depression” (2016) higher Hue (bluer) lower Saturation (grayer) lower Brightness (darker) 인스타그램으로 당신이 우울한지 알 수 있을까?
- 142. Rao (MVR) (24) . Results Both Alldata and Prediagnosis models were decisively superior to a null model . Alldata predictors were significant with 99% probability.57.5;(KAll =1 K 49.8) Pre = 1 7 Prediagnosis and Alldata confidence levels were largely identical, with two exceptions: Prediagnosis Brightness decreased to 90% confidence, and Prediagnosis posting frequency dropped to 30% confidence, suggesting a null predictive value in the latter case. Increased hue, along with decreased brightness and saturation, predicted depression. This means that photos posted by depressed individuals tended to be bluer, darker, and grayer (see Fig. 2). The more comments Instagram posts received, the more likely they were posted by depressed participants, but the opposite was true for likes received. In the Alldata model, higher posting frequency was also associated with depression. Depressed participants were more likely to post photos with faces, but had a lower average face count per photograph than healthy participants. Finally, depressed participants were less likely to apply Instagram filters to their posted photos. Fig. 2. Magnitude and direction of regression coefficients in Alldata (N=24,713) and Prediagnosis (N=18,513) models. Xaxis values represent the adjustment in odds of an observation belonging to depressed individuals, per Reece & Danforth, “Instagram photos reveal predictive markers of depression” (2016) Fig. 1. Comparison of HSV values. Right photograph has higher Hue (bluer), lower Saturation (grayer), and lower Brightness (darker) than left photograph. Instagram photos posted by depressed individuals had HSV values shifted towards those in the right photograph, compared with photos posted by healthy individuals. Units of observation In determining the best time span for this analysis, we encountered a difficult question: When and for how long does depression occur? A diagnosis of depression does not indicate the persistence of a depressive state for every moment of every day, and to conduct analysis using an individual’s entire posting history as a single unit of observation is therefore rather specious. At the other extreme, to take each individual photograph as units of observation runs the risk of being too granular. DeChoudhury et al. (5) looked at all of a given user’s posts in a single day, and aggregated those data into perperson, perday units of observation. We adopted this precedent of “userdays” as a unit of analysis . 5 Statistical framework We used Bayesian logistic regression with uninformative priors to determine the strength of individual predictors. Two separate models were trained. The Alldata model used all collected data to address Hypothesis 1. The Prediagnosis model used all data collected from higher Hue (bluer) lower Saturation (grayer) lower Brightness (darker) Digital Phenotype: Your Instagram knows if you are depressed
- 143. Digital Phenotype: Your Instagramknows if you are depressed Reece & Danforth, “Instagram photos reveal predictive markers of depression” (2016) . In particular, depressedχ2 07.84, p .17e 64;( All = 9 = 9 − 1 13.80, p .87e 44)χ2Pre = 8 = 2 − 1 participants were less likely than healthy participants to use any filters at all. When depressed participants did employ filters, they most disproportionately favored the “Inkwell” filter, which converts color photographs to blackandwhite images. Conversely, healthy participants most disproportionately favored the Valencia filter, which lightens the tint of photos. Examples of filtered photographs are provided in SI Appendix VIII. Fig. 3. Instagram filter usage among depressed and healthy participants. Bars indicate difference between observed and expected usage frequencies, based on a Chisquared analysis of independence. Blue bars indicate disproportionate use of a filter by depressed compared to healthy participants, orange bars indicate the reverse.
- 144. Digital Phenotype: Your Instagramknows if you are depressed Reece & Danforth, “Instagram photos reveal predictive markers of depression” (2016) VIII. Instagram filter examples Fig. S8. Examples of Inkwell and Valencia Instagram filters. Inkwell converts color photos to blackandwhite, Valencia lightens tint. Depressed participants most favored Inkwell compared to healthy participants, Healthy participants
- 145. ‘Facebook for Patients’,PatientsLikeMe.com
- 146. ‘Facebook for Patients’,PatientsLikeMe.com Stephen Heywood Benjamin Heywood James Heywood Jeff Cole • In 2004, three MIT engineers established the service for their own brother who was suffered from ALS. • Until 2011, only patients of 22 chronic disease, including ALS, HIV, Parkinson’s.
- 147.
- 148. • Age/sex • Medicationhistory • E-mail When joining in PatientsLikeMe
- 149. Users can findand friends with patients like them, based on disease, stage, age, sex ... Finding Patients Like Me!
- 150. Patines can keeptheir medical journals in the ‘Wall’, recording conditions, treatments, symptoms… (They don’t have to lie, because it’s totally anonymous)
- 151. Medications he/she took ‘RealWorld’ Feedback from the Patients • How long he/she took the medication • Purpose for which he/she took the medication • Dose of the medication • Efficacy / side-effect of the medication
- 152.
- 153. X 10,000 personal journalpersonal journal personal journal personal journal personal journal personal journal personal journal personal journal Big Medical Data
- 154. Business Model ofPatientsLikeMe Sell the real world data of anonymous patients To pharmaceutical or insurace companies
- 155. 110,000+ adverse eventreports, on 1,000 different medications
- 156. •PatientsLikeMe의 모든 데이터를 Genentech과 5년간 공유하기로 계약 •과거에도 Sanofi Aventis, Merck 와 임상시험 환자 모집 등을 제휴
- 157. “FDA will assessthe platform’s feasibility as a way to generate adverse event reports, which the FDA uses to regulate drugs after their release into the market.” 2015.6.15
- 158. Lexapro (escitalopram) selective serotoninreuptake inhibitor (SSRI)
- 161. The main sideeffect reported by PatientsLikeMe users on selective serotonin reuptake inhibitor (SSRI) Lexapro (escitalopram) was “Decreased sex drive (libido),” at 24% (n = 149), whereas the clinical trial data on Lexapro report 3% (n = 715) Nat Biotech 2009 Brownstein et al. http://www.nature.com/nbt/journal/v27/n10/full/nbt1009-888.html#close
- 163. Step1. Measure theData • With your smartphone • With wearable devices (connected to smartphone) • Personal genome analysis • Social Media ... without even going to the hospital!
- 164. Step 2. Collectthe Data
- 166. Sci Transl Med2015
- 168.
- 169.
- 170. Epic MyChart EpicEHR Dexcom CGM Patients/User Devices EHR Hospital Whitings + Apple Watch Apps HealthKit
- 172. • 애플 HealthKit가 미국의 23개 선도병원 중에, 14개의 병원과 협력 • 경쟁 플랫폼 Google Fit, S-Health 보다 현저히 빠른 움직임 • Beth Israel Deaconess 의 CIO • “25만명의 환자들 중 상당수가 웨어러블로 각종 데이터 생산 중. 이 모든 디바이스에 인터페이스를 우리 병원은 제공할 수 없다. 하지만 애플이라면 가능하다.” 2015.2.5
- 173. • 버릴리(구글)의 베이스라인프로젝트 • 건강과 질병을 새롭게 정의하기 위한 프로젝트 • 4년 동안 만 명의 개인의 건강 상태를 면밀하게 추적하여 데이터를 축적 • 심박수와 수면패턴 및 유전 정보, 감정 상태, 진료기록, 가족력, 소변/타액/혈액 검사 등
- 175. NatureAmerica,Inc.,partofSpringerNature.Allrightsreserved. Intro a b Round 1 Coachingsessions Round 2 Coaching sessions Round 3 Coaching sessions Month 1 Month 2 Month 3 Month 4 Month 5 Month 6 Month 7 Month 8 Month 9 Clinical labs Cardiovascular HDL/LDL cholesterol, triglycerides, particle profiles, and other markers Blood sample Metabolomics Xenobiotics and metabolism-related small molecules Blood sample Diabetes risk Fasting glucose, HbA1c, insulin, and other markers Blood sample Inflammation IL-6, IL-8, and other markers Blood sample Nutrition and toxins Ferritin, vitamin D, glutathione, mercury, lead, and other markers Blood sample Genetics Whole genome sequence Blood sample Proteomics Inflammation, cardiovascular, liver, brain, and heart-related proteins Blood sample Gut microbiome 16S rRNA sequencing Stool sample Quantified self Daily activity Activity tracker Stress Four-point cortisol Saliva Nature Biotechnology 2017
- 176. Pioneer 100 WellnessProject • 108 individual • for 9 months, at 3-month interval • whole genome sequences • clinical tests • metabolites • proteomes • microbiomes • frequent measurement • activity (fitbit) actionable possibilities behavioural coaching (pilot of 100K person wellness project)
- 177. ©2017NatureAmerica,Inc.,partofSpringerNature.Allrightsreserved. Proteomics Genetic traits Microbiome Coriobacteriia Allergic sensitization GH NEMO CD40L REN TPA HSP 27 LEP SIRT2 IL 6 FABP4 IL 1RA EGF VEGF A CSTB BETA NGF PPBP(2) PPBP NCF2 4E BP1 STAM PB SIRT2 CSF 1IL 6 FGF 21 IL 10RA IL 18R1IL8IL7 TNFSF14 CCL20 FLT3L CXCL10CD5HGFAXIN1 VEGFAOPGDNEROSM APCSINHBCCRP(2)CRPCFHR1HGFAC MBL2 SERPINC1 GC PTGDS ACTA2 ACTA2(2) PDGF SUBUNIT B Deletion Cfhr1 Inflammatory Bowel Disease Activated Partial Thromboplastin Time Bladder Cancer Bilirubin Levels Gamma Linolenic Acid Dihomo gamma Linolenic Acid Arachidonic Acid Linoleic Acid Adrenic Acid Deltaproteobacteria Mollicutes Verrucomicrobiae Coriobacteriales Verrucomicrobiales Verrucomicrobia Coriobacteriaceae 91otu13421 91otu4418 91otu1825 M ogibacteriaceae Unclassified Desulfovibrionaceae Pasteurellaceae Peptostreptococcaceae Christensenellaceae Verrucom icrobiaceae Alanine RatioOm6Om3 AlphaAminoN ButyricAcid Interleukinll6 SmallLdlParticle RatioGlnGln Threonine 3Methylhistidine AverageinflammationScore Mercury DocosapentaenoicAcidDocosatetraenoicAcid EicosadienoicAcidHomalrLeucineOmega3indexTyrosine HdlCholesterolCPeptide 1Methylhistidine 3HydroxyisovalericAcid IsovalerylglycineIsoleucine Figlu TotalCholesterolLinoleicDihomoYLinolejc PalmitoleicAcid ArachidonicAcid LdlParticle ArachidonicEicosapentaenoic Pasteurellales Diversity Tenericutes Clinical labs Metabolomics 5Hydroxyhexanoate Tl16:0(palmiticAcid) Tl18:3n6(gLinolenicAcid)Tl15:0(pentadecanoicAcid)Tl14:1n5(myristoleicAcid)Tl20:2n6(eicosadienoicAcid)Tl20:5n3(eicosapentaenoicAcid) Tl18:2n6(linoleicAcid) Tldm16:0(plasmalogenPalmiticAcid) Tl22:6n3(docosahexaenoicAcid) Tl22:4n6(adrenicAcid) Tl18:1n9(oleicAcid) Tldm18:1n9(plasmalogenOleicAcid) Tl20:4n6(arachidonicAcid) Tl14:0(myristicAcid) Arachidate(20:0) StearoylArachidonoylGlycerophosphoethanolamine(1)* 1Linoleoylglycerophosphocholine(18:2n6) StearoylLinoleoylGlycerophosphoethanolamine(1)* 1Palmitoleoylglycerophosphocholine(16:1)* PalmitoylOleoylGlycerophosphoglycerol(2)* PalmitoylLinoleoylGlycerophosphocholine(1)* Tl20:3n6(diHomoGLinoleicAcid) 2Hydroxypalmitate NervonoylSphingomyelin* Titl(totalTotalLipid) Cholesterol D ocosahexaenoate (dha;22;6n3) Eicosapentaenoate (epa; 20:5n3) 3 Carboxy 4 M ethyl 5 Propyl 2 Furanpropanoate (cm pf) 3 M ethyladipate Cholate Phosphoethanolamine 1 Oleoylglycerol (1 Monoolein) Tigloylglycine Valine sobutyrylglycine soleucine eucine P Cresol Glucuronide* Phenylacetylglutamine P Cresol Sulfate Tyrosine S Methylcysteine Cystine 3 Methylhistidine 1 Methylhistidine N Acetyltryptophan 3 Indoxyl Sulfate Serotonin (5ht) Creatinine Glutamate Cysteine Glutathione Disulfide Gamma Glutamylthreonine*Gamma Glutamylalanine Gamma Glutamylglutamate Gamma Glutamylglutamine Bradykinin, Hydroxy Pro(3) Bradykinin, Des Arg(9) BradykininMannoseBilirubin (e,e)* Biliverdin Bilirubin (z,z) L UrobilinNicotinamide Alpha TocopherolHippurate Cinnam oylglycine Ldl Particle N um ber Triglycerides Bilirubin Direct Alkaline Phosphatase EgfrNon AfrAm erican CholesterolTotal LdlSm all LdlM edium BilirubinTotal Ggt EgfrAfricanAmerican Cystine MargaricAcid ElaidicAcid Proinsulin Hba1c Insulin Triglycerides Ldlcholesterol DihomoGammaLinolenicAcid HsCrp GlutamicAcid Height Weight Leptin BodyMasIndex PhenylaceticAcid Valine TotalOmega3 TotalOmega6 HsCrpRelativeRisk DocosahexaenoicAcid AlphaAminoadipicAcid EicosapentaenoicAcid GammaAminobutyricAcid 5 Acetylam ino 6 Form ylam ino 3 M ethyluracil Adenosine 5 Monophosphate (amp) Gamma Glutamyltyrosine Gamma Glutamyl 2 Aminobutyrate N Acetyl 3 Methylhistidine* 3 Phenylpropionate (hydrocinnamate) Figure 2 Top 100 correlations per pair of data types. Subset of top statistically significant Spearman inter-omic cross-sectional correlations between all data sets collected in our cohort. Each line represents one correlation that was significant after adjustment for multiple hypothesis testing using the method of Benjamini and Hochberg10 at padj < 0.05. The mean of all three time points was used to compute the correlations between analytes. Up to 100 correlations per pair of data types are shown in this figure. See Supplementary Figure 1 and Supplementary Table 2 for the complete inter-omic cross-sectional network. Nature Biotechnology 2017 Top 100 correlations per pair of data types. Subset of top statistically significant Spearman inter-omic cross-sectional correlations between all data sets collected in our cohort.
- 178. •iCarbonX •중국 BGI의 대표였던준왕이 창업 •'모든 데이터를 측정'하고 이를 정밀 의료에 활용할 계획 •데이터를 측정할 수 있는 역량을 가진 회사에 투자 및 인수 •SomaLogic, HealthTell, PatientsLikMe •향후 5년 동안 100만명-1000만 명의 데이터 모을 계획 •이 데이터의 분석은 인공지능으로
- 179. Step 3. Insightfrom the Data
- 181.
- 182. How to Analyzeand Interpret the Big Data?
- 183. and/or Two ways toget insights from the big data
- 185. No choice butto bring AI into the medicine
- 186. Martin Duggan,“IBM WatsonHealth - Integrated Care & the Evolution to Cognitive Computing”
- 188. Jeopardy! 2011년 인간 챔피언두 명 과 퀴즈 대결을 벌여서 압도적인 우승을 차지
- 189. 600,000 pieces ofmedical evidence 2 million pages of text from 42 medical journals and clinical trials 69 guidelines, 61,540 clinical trials IBM Watson on Medicine Watson learned... + 1,500 lung cancer cases physician notes, lab results and clinical research + 14,700 hours of hands-on training
- 194. Annals of Oncology(2016) 27 (suppl_9): ix179-ix180. 10.1093/annonc/mdw601 Validation study to assess performance of IBM cognitive computing system Watson for oncology with Manipal multidisciplinary tumour board for 1000 consecutive cases: An Indian experience • MMDT(Manipal multidisciplinary tumour board) treatment recommendation and data of 1000 cases of 4 different cancers breast (638), colon (126), rectum (124) and lung (112) which were treated in last 3 years was collected. • Of the treatment recommendations given by MMDT, WFO provided 50% in REC, 28% in FC, 17% in NREC • Nearly 80% of the recommendations were in WFO REC and FC group • 5% of the treatment provided by MMDT was not available with WFO • The degree of concordance varied depending on the type of cancer • WFO-REC was high in Rectum (85%) and least in Lung (17.8%) • high with TNBC (67.9%); HER2 negative (35%) • WFO took a median of 40 sec to capture, analyze and give the treatment. (vs MMDT took the median time of 15 min)
- 195. WFO in ASCO2017 • Early experience with IBM WFO cognitive computing system for lung and colorectal cancer treatment (마니팔 병원) • 지난 3년간: lung cancer(112), colon cancer(126), rectum cancer(124) • lung cancer: localized 88.9%, meta 97.9% • colon cancer: localized 85.5%, meta 76.6% • rectum cancer: localized 96.8%, meta 80.6% Performance of WFO in India 2017 ASCO annual Meeting, J Clin Oncol 35, 2017 (suppl; abstr 8527)
- 196. San Antonio BreastCancer Symposium—December 6-10, 2016 Concordance WFO (@T2) and MMDT (@T1* v. T2**) (N= 638 Breast Cancer Cases) Time Point /Concordance REC REC + FC n % n % T1* 296 46 463 73 T2** 381 60 574 90 This presentation is the intellectual property of the author/presenter.Contact somusp@yahoo.com for permission to reprint and/or distribute.26 * T1 Time of original treatment decision by MMDT in the past (last 1-3 years) ** T2 Time (2016) of WFO’s treatment advice and of MMDT’s treatment decision upon blinded re-review of non-concordant cases
- 197. 잠정적 결론 •왓슨 포온콜로지와 의사의 일치율: •암종별로 다르다. •같은 암종에서도 병기별로 다르다. •같은 암종에 대해서도 병원별/국가별로 다르다. •시간이 흐름에 따라 달라질 가능성이 있다.
- 198. 원칙이 필요하다 •어떤 환자의경우, 왓슨에게 의견을 물을 것인가? •왓슨을 (암종별로) 얼마나 신뢰할 것인가? •왓슨의 의견을 환자에게 공개할 것인가? •왓슨과 의료진의 판단이 다른 경우 어떻게 할 것인가? •왓슨에게 보험 급여를 매길 수 있는가? 이러한 기준에 따라 의료의 질/치료효과가 달라질 수 있으나, 현재 개별 병원이 개별적인 기준으로 활용하게 됨
- 199. Empowering the OncologyCommunity for Cancer Care Genomics Oncology Clinical Trial Matching Watson Health’s oncology clients span more than 35 hospital systems “Empowering the Oncology Community for Cancer Care” Andrew Norden, KOTRA Conference, March 2017, “The Future of Health is Cognitive”
- 200. IBM Watson Health Watsonfor Clinical Trial Matching (CTM) 18 1. According to the National Comprehensive Cancer Network (NCCN) 2. http://csdd.tufts.edu/files/uploads/02_-_jan_15,_2013_-_recruitment-retention.pdf© 2015 International Business Machines Corporation Searching across eligibility criteria of clinical trials is time consuming and labor intensive Current Challenges Fewer than 5% of adult cancer patients participate in clinical trials1 37% of sites fail to meet minimum enrollment targets. 11% of sites fail to enroll a single patient 2 The Watson solution • Uses structured and unstructured patient data to quickly check eligibility across relevant clinical trials • Provides eligible trial considerations ranked by relevance • Increases speed to qualify patients Clinical Investigators (Opportunity) • Trials to Patient: Perform feasibility analysis for a trial • Identify sites with most potential for patient enrollment • Optimize inclusion/exclusion criteria in protocols Faster, more efficient recruitment strategies, better designed protocols Point of Care (Offering) • Patient to Trials: Quickly find the right trial that a patient might be eligible for amongst 100s of open trials available Improve patient care quality, consistency, increased efficiencyIBM Confidential
- 201. •총 16주간 HOG(Highlands Oncology Group)의 폐암과 유방암 환자 2,620명을 대상 •90명의 환자를 3개의 노바티스 유방암 임상 프로토콜에 따라 선별 •임상 시험 코디네이터: 1시간 50분 •Watson CTM: 24분 (78% 시간 단축) •Watson CTM은 임상 시험 기준에 해당되지 않는 환자 94%를 자동으로 스크리닝
- 202. Watson Genomics Overview 20 WatsonGenomics Content • 20+ Content Sources Including: • Medical Articles (23Million) • Drug Information • Clinical Trial Information • Genomic Information Case Sequenced VCF / MAF, Log2, Dge Encryption Molecular Profile Analysis Pathway Analysis Drug Analysis Service Analysis, Reports, & Visualizations
- 203.
- 204. DeepFace: Closing theGap to Human-Level Performance in FaceVerification Taigman,Y. et al. (2014). DeepFace: Closing the Gap to Human-Level Performance in FaceVerification, CVPR’14. Figure 2. Outline of the DeepFace architecture. A front-end of a single convolution-pooling-convolution filtering on the rectified input, followed by three locally-connected layers and two fully-connected layers. Colors illustrate feature maps produced at each layer. The net includes more than 120 million parameters, where more than 95% come from the local and fully connected layers. very few parameters. These layers merely expand the input into a set of simple local features. The subsequent layers (L4, L5 and L6) are instead lo- cally connected [13, 16], like a convolutional layer they ap- ply a filter bank, but every location in the feature map learns a different set of filters. Since different regions of an aligned image have different local statistics, the spatial stationarity The goal of training is to maximize the probability of the correct class (face id). We achieve this by minimiz- ing the cross-entropy loss for each training sample. If k is the index of the true label for a given input, the loss is: L = log pk. The loss is minimized over the parameters by computing the gradient of L w.r.t. the parameters and Human: 95% vs. DeepFace in Facebook: 97.35% Recognition Accuracy for Labeled Faces in the Wild (LFW) dataset (13,233 images, 5,749 people)
- 205. FaceNet:A Unified Embeddingfor Face Recognition and Clustering Schroff, F. et al. (2015). FaceNet:A Unified Embedding for Face Recognition and Clustering Human: 95% vs. FaceNet of Google: 99.63% Recognition Accuracy for Labeled Faces in the Wild (LFW) dataset (13,233 images, 5,749 people) False accept False reject s. This shows all pairs of images that were on LFW. Only eight of the 13 errors shown he other four are mislabeled in LFW. on Youtube Faces DB ge similarity of all pairs of the first one our face detector detects in each video. False accept False reject Figure 6. LFW errors. This shows all pairs of images that were incorrectly classified on LFW. Only eight of the 13 errors shown here are actual errors the other four are mislabeled in LFW. 5.7. Performance on Youtube Faces DB We use the average similarity of all pairs of the first one hundred frames that our face detector detects in each video. This gives us a classification accuracy of 95.12%±0.39. Using the first one thousand frames results in 95.18%. Compared to [17] 91.4% who also evaluate one hundred frames per video we reduce the error rate by almost half. DeepId2+ [15] achieved 93.2% and our method reduces this error by 30%, comparable to our improvement on LFW. 5.8. Face Clustering Our compact embedding lends itself to be used in order to cluster a users personal photos into groups of people with the same identity. The constraints in assignment imposed by clustering faces, compared to the pure verification task, lead to truly amazing results. Figure 7 shows one cluster in a users personal photo collection, generated using agglom- erative clustering. It is a clear showcase of the incredible invariance to occlusion, lighting, pose and even age. Figure 7. Face Clustering. Shown is an exemplar cluster for one user. All these images in the users personal photo collection were clustered together. 6. Summary We provide a method to directly learn an embedding into an Euclidean space for face verification. This sets it apart from other methods [15, 17] who use the CNN bottleneck layer, or require additional post-processing such as concate- nation of multiple models and PCA, as well as SVM clas- sification. Our end-to-end training both simplifies the setup and shows that directly optimizing a loss relevant to the task at hand improves performance. Another strength of our model is that it only requires False accept False reject Figure 6. LFW errors. This shows all pairs of images that were incorrectly classified on LFW. Only eight of the 13 errors shown here are actual errors the other four are mislabeled in LFW. 5.7. Performance on Youtube Faces DB We use the average similarity of all pairs of the first one hundred frames that our face detector detects in each video. This gives us a classification accuracy of 95.12%±0.39. Using the first one thousand frames results in 95.18%. Compared to [17] 91.4% who also evaluate one hundred frames per video we reduce the error rate by almost half. DeepId2+ [15] achieved 93.2% and our method reduces this error by 30%, comparable to our improvement on LFW. 5.8. Face Clustering Our compact embedding lends itself to be used in order to cluster a users personal photos into groups of people with the same identity. The constraints in assignment imposed by clustering faces, compared to the pure verification task, Figure 7. Face Clustering. Shown is an exemplar cluster for one user. All these images in the users personal photo collection were clustered together. 6. Summary We provide a method to directly learn an embedding into an Euclidean space for face verification. This sets it apart from other methods [15, 17] who use the CNN bottleneck layer, or require additional post-processing such as concate- nation of multiple models and PCA, as well as SVM clas-
- 206. Targeting Ultimate Accuracy:Face Recognition via Deep Embedding Jingtuo Liu (2015) Targeting Ultimate Accuracy: Face Recognition via Deep Embedding Human: 95% vs.Baidu: 99.77% Recognition Accuracy for Labeled Faces in the Wild (LFW) dataset (13,233 images, 5,749 people) 3 Although several algorithms have achieved nearly perfect accuracy in the 6000-pair verification task, a more practical can achieve 95.8% identification rate, relatively reducing the error rate by about 77%. TABLE 3. COMPARISONS WITH OTHER METHODS ON SEVERAL EVALUATION TASKS Score = -0.060 (pair #113) Score = -0.022 (pair #202) Score = -0.034 (pair #656) Score = -0.031 (pair #1230) Score = -0.073 (pair #1862) Score = -0.091(pair #2499) Score = -0.024 (pair #2551) Score = -0.036 (pair #2552) Score = -0.089 (pair #2610) Method Performance on tasks Pair-wise Accuracy(%) Rank-1(%) DIR(%) @ FAR =1% Verification(% )@ FAR=0.1% Open-set Identification(% )@ Rank = 1,FAR = 0.1% IDL Ensemble Model 99.77 98.03 95.8 99.41 92.09 IDL Single Model 99.68 97.60 94.12 99.11 89.08 FaceNet[12] 99.63 NA NA NA NA DeepID3[9] 99.53 96.00 81.40 NA NA Face++[2] 99.50 NA NA NA NA Facebook[15] 98.37 82.5 61.9 NA NA Learning from Scratch[4] 97.73 NA NA 80.26 28.90 HighDimLBP[10] 95.17 NA NA 41.66(reported in [4]) 18.07(reported in [4]) • 6,000쌍의 얼굴 사진 중에 바이두의 인공지능은 불과 14쌍만을 잘못 판단 • 알고 보니 이 14쌍 중의 5쌍의 사진은 오히려 정답에 오류가 있었고, 실제로는 인공지능이 정확 (red box)
- 207. Show and Tell: ANeural Image Caption Generator Vinyals, O. et al. (2015). Show and Tell:A Neural Image Caption Generator, arXiv:1411.4555 v om Samy Bengio Google bengio@google.com Dumitru Erhan Google dumitru@google.com s a cts his re- m- ed he de- nts A group of people shopping at an outdoor market. ! There are many vegetables at the fruit stand. Vision! Deep CNN Language ! Generating! RNN Figure 1. NIC, our model, is based end-to-end on a neural net- work consisting of a vision CNN followed by a language gener-
- 208. Show and Tell: ANeural Image Caption Generator Vinyals, O. et al. (2015). Show and Tell:A Neural Image Caption Generator, arXiv:1411.4555 Figure 5. A selection of evaluation results, grouped by human rating.
- 209.
- 210. Bone Age Assessment •M: 28 Classes • F: 20 Classes • Method: G.P. • Top3-95.28% (F) • Top3-81.55% (M)
- 211. Business Area Medical ImageAnalysis VUNOnet and our machine learning technology will help doctors and hospitals manage medical scans and images intelligently to make diagnosis faster and more accurately. Original Image Automatic Segmentation EmphysemaNormal ReticularOpacity Our system finds DILDs at the highest accuracy * DILDs: Diffuse Interstitial Lung Disease Digital Radiologist Collaboration with Prof. Joon Beom Seo (Asan Medical Center) Analysed 1200 patients for 3 months
- 212. Detection of DiabeticRetinopathy
- 213. 당뇨성 망막병증 • 당뇨병의대표적 합병증: 당뇨병력이 30년 이상 환자 90% 발병 • 안과 전문의들이 안저(안구의 안쪽)를 사진으로 찍어서 판독 • 망막 내 미세혈관 생성, 출혈, 삼출물 정도를 파악하여 진단
- 214. • EyePACS-1 과 Messidor-2의 AUC = 0.991, 0.990 • 7-8명의 안과 전문의와 sensitivity, specificity 가 동일한 수준 • F-score: 0.95 (vs. 인간 의사는 0.91) Additional sensitivity analyses were conducted for sev- eralsubcategories:(1)detectingmoderateorworsediabeticreti- effects of data set size on algorithm performance were exam- ined and shown to plateau at around 60 000 images (or ap- Figure 2. Validation Set Performance for Referable Diabetic Retinopathy 100 80 60 40 20 0 0 70 80 85 95 90 75 0 5 10 15 20 25 30 100806040 Sensitivity,% 1 – Specificity, % 20 EyePACS-1: AUC, 99.1%; 95% CI, 98.8%-99.3%A 100 High-sensitivity operating point High-specificity operating point 100 80 60 40 20 0 0 70 80 85 95 90 75 0 5 10 15 20 25 30 100806040 Sensitivity,% 1 – Specificity, % 20 Messidor-2: AUC, 99.0%; 95% CI, 98.6%-99.5%B 100 High-specificity operating point High-sensitivity operating point Performance of the algorithm (black curve) and ophthalmologists (colored circles) for the presence of referable diabetic retinopathy (moderate or worse diabetic retinopathy or referable diabetic macular edema) on A, EyePACS-1 (8788 fully gradable images) and B, Messidor-2 (1745 fully gradable images). The black diamonds on the graph correspond to the sensitivity and specificity of the algorithm at the high-sensitivity and high-specificity operating points. In A, for the high-sensitivity operating point, specificity was 93.4% (95% CI, 92.8%-94.0%) and sensitivity was 97.5% (95% CI, 95.8%-98.7%); for the high-specificity operating point, specificity was 98.1% (95% CI, 97.8%-98.5%) and sensitivity was 90.3% (95% CI, 87.5%-92.7%). In B, for the high-sensitivity operating point, specificity was 93.9% (95% CI, 92.4%-95.3%) and sensitivity was 96.1% (95% CI, 92.4%-98.3%); for the high-specificity operating point, specificity was 98.5% (95% CI, 97.7%-99.1%) and sensitivity was 87.0% (95% CI, 81.1%-91.0%). There were 8 ophthalmologists who graded EyePACS-1 and 7 ophthalmologists who graded Messidor-2. AUC indicates area under the receiver operating characteristic curve. Research Original Investigation Accuracy of a Deep Learning Algorithm for Detection of Diabetic Retinopathy Results
- 215.
- 216.
- 217. Skin cancer classificationperformance of the CNN and dermatologists. LETT a b 0 1 Sensitivity 0 1 Specificity Melanoma: 130 images 1 Specificity Melanoma: 225 images 0 1 Sensitivity 0 1 Specificity Melanoma: 111 dermoscopy images 1 Specificity Carcinoma: 707 images 1 Specificity Melanoma: 1,010 dermoscopy images 0 1 Sensitivity 0 1 Specificity Carcinoma: 135 images Algorithm: AUC = 0.96 Dermatologists (25) Average dermatologist Algorithm: AUC = 0.94 Dermatologists (22) Average dermatologist Algorithm: AUC = 0.91 Dermatologists (21) Average dermatologist 21명 중에 인공지능보다 정확성이 떨어지는 피부과 전문의들이 상당수 있었음 피부과 전문의들의 평균 성적도 인공지능보다 좋지 않았음
- 218.
- 219.
- 220.
- 221. WSJ, 2017 June •다국적 제약사는 인공지능 기술을 신약 개발에 활용하기 위해 다양한 시도 • 최근 인공지능에서는 과거의 virtual screening, docking 등과는 다른 방식을 이용
- 223. AtomNet: A DeepConvolutional Neural Network for Bioactivity Prediction in Structure-based Drug Discovery Izhar Wallach Atomwise, Inc. izhar@atomwise.com Michael Dzamba Atomwise, Inc. misko@atomwise.com Abraham Heifets Atomwise, Inc. abe@atomwise.com Abstract Deep convolutional neural networks comprise a subclass of deep neural networks (DNN) with a constrained architecture that leverages the spatial and temporal structure of the domain they model. Convolutional networks achieve the best pre- dictive performance in areas such as speech and image recognition by hierarchi- cally composing simple local features into complex models. Although DNNs have been used in drug discovery for QSAR and ligand-based bioactivity predictions, none of these models have benefited from this powerful convolutional architec- ture. This paper introduces AtomNet, the first structure-based, deep convolutional neural network designed to predict the bioactivity of small molecules for drug dis- covery applications. We demonstrate how to apply the convolutional concepts of feature locality and hierarchical composition to the modeling of bioactivity and chemical interactions. In further contrast to existing DNN techniques, we show that AtomNet’s application of local convolutional filters to structural target infor- mation successfully predicts new active molecules for targets with no previously known modulators. Finally, we show that AtomNet outperforms previous docking approaches on a diverse set of benchmarks by a large margin, achieving an AUC greater than 0.9 on 57.8% of the targets in the DUDE benchmark. 1 Introduction Fundamentally, biological systems operate through the physical interaction of molecules. The ability to determine when molecular binding occurs is therefore critical for the discovery of new medicines and for furthering of our understanding of biology. Unfortunately, despite thirty years of compu- tational efforts, computer tools remain too inaccurate for routine binding prediction, and physical experiments remain the state of the art for binding determination. The ability to accurately pre- dict molecular binding would reduce the time-to-discovery of new treatments, help eliminate toxic molecules early in development, and guide medicinal chemistry efforts [1, 2]. In this paper, we introduce a new predictive architecture, AtomNet, to help address these challenges. AtomNet is novel in two regards: AtomNet is the first deep convolutional neural network for molec- ular binding affinity prediction. It is also the first deep learning system that incorporates structural information about the target to make its predictions. Deep convolutional neural networks (DCNN) are currently the best performing predictive models for speech and vision [3, 4, 5, 6]. DCNN is a class of deep neural network that constrains its model architecture to leverage the spatial and temporal structure of its domain. For example, a low-level image feature, such as an edge, can be described within a small spatially-proximate patch of pixels. Such a feature detector can share evidence across the entire receptive field by “tying the weights” of the detector neurons, as the recognition of the edge does not depend on where it is found within 1 arXiv:1510.02855v1[cs.LG]10Oct2015
- 224. AtomNet: A DeepConvolutional Neural Network for Bioactivity Prediction in Structure-based Drug Discovery Izhar Wallach Atomwise, Inc. izhar@atomwise.com Michael Dzamba Atomwise, Inc. misko@atomwise.com Abraham Heifets Atomwise, Inc. abe@atomwise.com Abstract Deep convolutional neural networks comprise a subclass of deep neural networks (DNN) with a constrained architecture that leverages the spatial and temporal structure of the domain they model. Convolutional networks achieve the best pre- dictive performance in areas such as speech and image recognition by hierarchi- cally composing simple local features into complex models. Although DNNs have been used in drug discovery for QSAR and ligand-based bioactivity predictions, none of these models have benefited from this powerful convolutional architec- ture. This paper introduces AtomNet, the first structure-based, deep convolutional neural network designed to predict the bioactivity of small molecules for drug dis- covery applications. We demonstrate how to apply the convolutional concepts of feature locality and hierarchical composition to the modeling of bioactivity and chemical interactions. In further contrast to existing DNN techniques, we show that AtomNet’s application of local convolutional filters to structural target infor- mation successfully predicts new active molecules for targets with no previously known modulators. Finally, we show that AtomNet outperforms previous docking approaches on a diverse set of benchmarks by a large margin, achieving an AUC greater than 0.9 on 57.8% of the targets in the DUDE benchmark. 1 Introduction Fundamentally, biological systems operate through the physical interaction of molecules. The ability to determine when molecular binding occurs is therefore critical for the discovery of new medicines and for furthering of our understanding of biology. Unfortunately, despite thirty years of compu- tational efforts, computer tools remain too inaccurate for routine binding prediction, and physical experiments remain the state of the art for binding determination. The ability to accurately pre- dict molecular binding would reduce the time-to-discovery of new treatments, help eliminate toxic molecules early in development, and guide medicinal chemistry efforts [1, 2]. In this paper, we introduce a new predictive architecture, AtomNet, to help address these challenges. AtomNet is novel in two regards: AtomNet is the first deep convolutional neural network for molec- ular binding affinity prediction. It is also the first deep learning system that incorporates structural information about the target to make its predictions. Deep convolutional neural networks (DCNN) are currently the best performing predictive models for speech and vision [3, 4, 5, 6]. DCNN is a class of deep neural network that constrains its model architecture to leverage the spatial and temporal structure of its domain. For example, a low-level image feature, such as an edge, can be described within a small spatially-proximate patch of pixels. Such a feature detector can share evidence across the entire receptive field by “tying the weights” of the detector neurons, as the recognition of the edge does not depend on where it is found within 1 arXiv:1510.02855v1[cs.LG]10Oct2015 Smina 123 35 5 0 0 Table 3: The number of targets on which AtomNet and Smina exceed given adjusted-logAUC thresh- olds. For example, on the CHEMBL-20 PMD set, AtomNet achieves an adjusted-logAUC of 0.3 or better for 27 targets (out of 50 possible targets). ChEMBL-20 PMD contains 50 targets, DUDE- 30 contains 30 targets, DUDE-102 contains 102 targets, and ChEMBL-20 inactives contains 149 targets. To overcome these limitations we take an indirect approach. Instead of directly visualizing filters in order to understand their specialization, we apply filters to input data and examine the location where they maximally fire. Using this technique we were able to map filters to chemical functions. For example, Figure 5 illustrate the 3D locations at which a particular filter from our first convo- lutional layer fires. Visual inspection of the locations at which that filter is active reveals that this filter specializes as a sulfonyl/sulfonamide detector. This demonstrates the ability of the model to learn complex chemical features from simpler ones. In this case, the filter has inferred a meaningful spatial arrangement of input atom types without any chemical prior knowledge. Figure 5: Sulfonyl/sulfonamide detection with autonomously trained convolutional filters. 8 • 이미 알려진 단백질-리간드 3차원 결합 구조를 딥러닝(CNN)으로 학습 • 화학 결합 등에 대한 계산 없이도, 단백질-리간드 결합 여부를 계산 • 기존의 구조기반 예측 등 대비, 딥러닝으로 더 정확히 예측하였음
- 225. AtomNet: A DeepConvolutional Neural Network for Bioactivity Prediction in Structure-based Drug Discovery Izhar Wallach Atomwise, Inc. izhar@atomwise.com Michael Dzamba Atomwise, Inc. misko@atomwise.com Abraham Heifets Atomwise, Inc. abe@atomwise.com Abstract Deep convolutional neural networks comprise a subclass of deep neural networks (DNN) with a constrained architecture that leverages the spatial and temporal structure of the domain they model. Convolutional networks achieve the best pre- dictive performance in areas such as speech and image recognition by hierarchi- cally composing simple local features into complex models. Although DNNs have been used in drug discovery for QSAR and ligand-based bioactivity predictions, none of these models have benefited from this powerful convolutional architec- ture. This paper introduces AtomNet, the first structure-based, deep convolutional neural network designed to predict the bioactivity of small molecules for drug dis- covery applications. We demonstrate how to apply the convolutional concepts of feature locality and hierarchical composition to the modeling of bioactivity and chemical interactions. In further contrast to existing DNN techniques, we show that AtomNet’s application of local convolutional filters to structural target infor- mation successfully predicts new active molecules for targets with no previously known modulators. Finally, we show that AtomNet outperforms previous docking approaches on a diverse set of benchmarks by a large margin, achieving an AUC greater than 0.9 on 57.8% of the targets in the DUDE benchmark. 1 Introduction Fundamentally, biological systems operate through the physical interaction of molecules. The ability to determine when molecular binding occurs is therefore critical for the discovery of new medicines and for furthering of our understanding of biology. Unfortunately, despite thirty years of compu- tational efforts, computer tools remain too inaccurate for routine binding prediction, and physical experiments remain the state of the art for binding determination. The ability to accurately pre- dict molecular binding would reduce the time-to-discovery of new treatments, help eliminate toxic molecules early in development, and guide medicinal chemistry efforts [1, 2]. In this paper, we introduce a new predictive architecture, AtomNet, to help address these challenges. AtomNet is novel in two regards: AtomNet is the first deep convolutional neural network for molec- ular binding affinity prediction. It is also the first deep learning system that incorporates structural information about the target to make its predictions. Deep convolutional neural networks (DCNN) are currently the best performing predictive models for speech and vision [3, 4, 5, 6]. DCNN is a class of deep neural network that constrains its model architecture to leverage the spatial and temporal structure of its domain. For example, a low-level image feature, such as an edge, can be described within a small spatially-proximate patch of pixels. Such a feature detector can share evidence across the entire receptive field by “tying the weights” of the detector neurons, as the recognition of the edge does not depend on where it is found within 1 arXiv:1510.02855v1[cs.LG]10Oct2015 • 이미 알려진 단백질-리간드 3차원 결합 구조를 딥러닝(CNN)으로 학습 • 화학 결합 등에 대한 계산 없이도, 단백질-리간드 결합 여부를 계산 • 기존의 구조기반 예측 등 대비, 딥러닝으로 더 정확히 예측하였음
- 227. 604 VOLUME 35NUMBER 7 JULY 2017 NATURE BIOTECHNOLOGY AI-powered drug discovery captures pharma interest Adrug-huntingdealinkedlastmonth,between Numerate,ofSanBruno,California,andTakeda PharmaceuticaltouseNumerate’sartificialintel- ligence (AI) suite to discover small-molecule therapies for oncology, gastroenterology and central nervous system disorders, is the latest in a growing number of research alliances involv- ing AI-powered computational drug develop- ment firms. Also last month, GNS Healthcare of Cambridge, Massachusetts announced a deal with Roche subsidiary Genentech of South San Francisco, California to use GNS’s AI platform to better understand what affects the efficacy of knowntherapiesinoncology.InMay,Exscientia of Dundee, Scotland, signed a deal with Paris- based Sanofi that includes up to €250 ($280) million in milestone payments. Exscientia will provide the compound design and Sanofi the chemical synthesis of new drugs for diabetes and cardiovascular disease. The trend indicates thatthepharmaindustry’slong-runningskepti- cism about AI is softening into genuine interest, driven by AI’s promise to address the industry’s principal pain point: clinical failure rates. The industry’s willingness to consider AI approaches reflects the reality that drug discov- eryislaborious,timeconsumingandnotpartic- ularly effective. A two-decade-long downward trend in clinical success rates has only recently improved (Nat. Rev. Drug Disc. 15, 379–380, 2016). Still, today, only about one in ten drugs thatenterphase1clinicaltrialsreachespatients. Half those failures are due to a lack of efficacy, says Jackie Hunter, CEO of BenevolentBio, a division of BenevolentAI of London. “That tells you we’re not picking the right targets,” she says. “Even a 5 or 10% reduction in efficacy failure would be amazing.” Hunter’s views on AI in drug discovery are featured in Ernst & Young’s BiotechnologyReport2017releasedlastmonth. Companies that have been watching AI from the sidelines are now jumping in. The best- known machine-learning model for drug dis- covery is perhaps IBM’s Watson. IBM signed a deal in December 2016 with Pfizer to aid the pharma giant’s immuno-oncology drug discov- eryefforts,addingtoastringofpreviousdealsin the biopharma space (Nat.Biotechnol.33,1219– 1220, 2015). IBM’s Watson hunts for drugs by sorting through vast amounts of textual data to provide quick analyses, and tests hypotheses by sorting through massive amounts of laboratory data, clinical reports and scientific publications. BenevolentAI takes a similar approach with algorithms that mine the research literature and proprietary research databases. The explosion of biomedical data has driven much of industry’s interest in AI (Table 1). The confluence of ever-increasing computational horsepower and the proliferation of large data sets has prompted scientists to seek learning algorithms that can help them navigate such massive volumes of information. A lot of the excitement about AI in drug discovery has spilled over from other fields. Machine vision, which allows, among other things, self-driving cars, and language process- ing have given rise to sophisticated multilevel artificial neural networks known as deep- learning algorithms that can be used to model biological processes from assay data as well as textual data. In the past people didn’t have enough data to properly train deep-learning algorithms, says Mark Gerstein, a biomedical informat- ics professor at Yale University in New Haven, Connecticut.Nowresearchershavebeenableto build massive databases and harness them with these algorithms, he says. “I think that excite- ment is justified.” Numerate is one of a growing number of AI companies founded to take advantage of that dataonslaughtasappliedtodrugdiscovery.“We apply AI to chemical design at every stage,” says Guido Lanza, Numerate’s CEO. It will provide Tokyo-basedTakedawithcandidatesforclinical trials by virtual compound screenings against targets, designing and optimizing compounds, andmodelingabsorption,distribution,metabo- lism and excretion, and toxicity. The agreement includes undisclosed milestone payments and royalties. Academic laboratories are also embracing AI tools. In April, Atomwise of San Francisco launched its Artificial Intelligence Molecular Screen awards program, which will deliver 72 potentially therapeutic compounds to as many as 100 university research labs at no charge. Atomwise is a University of Toronto spinout that in 2015 secured an alliance with Merck of Kenilworth, New Jersey. For this new endeavor, it will screen 10 million molecules using its AtomNet platform to provide each lab with 72 compounds aimed at a specific target of the laboratory’s choosing. The Japanese government launched in 2016 a research consortium centered on using Japan’s K supercomputer to ramp up drug discovery efficiency across dozens of local companies and institutions. Among those involved are Takeda and tech giants Fujitsu of Tokyo, Japan, and NEC, also of Tokyo, as well as Kyoto University Hospital and Riken, Japan’s National Research and Development Institute, which will provide clinical data. Deep learning is starting to gain acolytes in the drug discovery space. KTSDESIGN/SciencePhotoLibrary N E W S©2017NatureAmerica,Inc.,partofSpringerNature.Allrightsreserved.
- 228. 604 VOLUME 35NUMBER 7 JULY 2017 NATURE BIOTECHNOLOGY AI-powered drug discovery captures pharma interest Adrug-huntingdealinkedlastmonth,between Numerate,ofSanBruno,California,andTakeda PharmaceuticaltouseNumerate’sartificialintel- ligence (AI) suite to discover small-molecule therapies for oncology, gastroenterology and central nervous system disorders, is the latest in a growing number of research alliances involv- ing AI-powered computational drug develop- ment firms. Also last month, GNS Healthcare of Cambridge, Massachusetts announced a deal with Roche subsidiary Genentech of South San Francisco, California to use GNS’s AI platform to better understand what affects the efficacy of knowntherapiesinoncology.InMay,Exscientia of Dundee, Scotland, signed a deal with Paris- based Sanofi that includes up to €250 ($280) million in milestone payments. Exscientia will provide the compound design and Sanofi the chemical synthesis of new drugs for diabetes and cardiovascular disease. The trend indicates thatthepharmaindustry’slong-runningskepti- cism about AI is softening into genuine interest, driven by AI’s promise to address the industry’s principal pain point: clinical failure rates. The industry’s willingness to consider AI approaches reflects the reality that drug discov- eryislaborious,timeconsumingandnotpartic- ularly effective. A two-decade-long downward trend in clinical success rates has only recently improved (Nat. Rev. Drug Disc. 15, 379–380, 2016). Still, today, only about one in ten drugs thatenterphase1clinicaltrialsreachespatients. Half those failures are due to a lack of efficacy, says Jackie Hunter, CEO of BenevolentBio, a division of BenevolentAI of London. “That tells you we’re not picking the right targets,” she says. “Even a 5 or 10% reduction in efficacy failure would be amazing.” Hunter’s views on AI in drug discovery are featured in Ernst & Young’s BiotechnologyReport2017releasedlastmonth. Companies that have been watching AI from the sidelines are now jumping in. The best- known machine-learning model for drug dis- covery is perhaps IBM’s Watson. IBM signed a deal in December 2016 with Pfizer to aid the pharma giant’s immuno-oncology drug discov- eryefforts,addingtoastringofpreviousdealsin the biopharma space (Nat.Biotechnol.33,1219– 1220, 2015). IBM’s Watson hunts for drugs by sorting through vast amounts of textual data to provide quick analyses, and tests hypotheses by sorting through massive amounts of laboratory data, clinical reports and scientific publications. BenevolentAI takes a similar approach with algorithms that mine the research literature and proprietary research databases. The explosion of biomedical data has driven much of industry’s interest in AI (Table 1). The confluence of ever-increasing computational horsepower and the proliferation of large data sets has prompted scientists to seek learning algorithms that can help them navigate such massive volumes of information. A lot of the excitement about AI in drug discovery has spilled over from other fields. Machine vision, which allows, among other things, self-driving cars, and language process- ing have given rise to sophisticated multilevel artificial neural networks known as deep- learning algorithms that can be used to model biological processes from assay data as well as textual data. In the past people didn’t have enough data to properly train deep-learning algorithms, says Mark Gerstein, a biomedical informat- ics professor at Yale University in New Haven, Connecticut.Nowresearchershavebeenableto build massive databases and harness them with these algorithms, he says. “I think that excite- ment is justified.” Numerate is one of a growing number of AI companies founded to take advantage of that dataonslaughtasappliedtodrugdiscovery.“We apply AI to chemical design at every stage,” says Guido Lanza, Numerate’s CEO. It will provide Tokyo-basedTakedawithcandidatesforclinical trials by virtual compound screenings against targets, designing and optimizing compounds, andmodelingabsorption,distribution,metabo- lism and excretion, and toxicity. The agreement includes undisclosed milestone payments and royalties. Academic laboratories are also embracing AI tools. In April, Atomwise of San Francisco launched its Artificial Intelligence Molecular Screen awards program, which will deliver 72 potentially therapeutic compounds to as many as 100 university research labs at no charge. Atomwise is a University of Toronto spinout that in 2015 secured an alliance with Merck of Kenilworth, New Jersey. For this new endeavor, it will screen 10 million molecules using its AtomNet platform to provide each lab with 72 compounds aimed at a specific target of the laboratory’s choosing. The Japanese government launched in 2016 a research consortium centered on using Japan’s K supercomputer to ramp up drug discovery efficiency across dozens of local companies and institutions. Among those involved are Takeda and tech giants Fujitsu of Tokyo, Japan, and NEC, also of Tokyo, as well as Kyoto University Hospital and Riken, Japan’s National Research and Development Institute, which will provide clinical data. Deep learning is starting to gain acolytes in the drug discovery space. KTSDESIGN/SciencePhotoLibrary N E W S©2017NatureAmerica,Inc.,partofSpringerNature.Allrightsreserved. Genomics data analytics startup WuXi NextCode Genomics of Shanghai; Cambridge, Massachusetts; and Reykjavík, Iceland, collab- orated with researchers at Yale University on a study that used the company’s deep-learning algorithm to identify a key mechanism in blood vessel growth. The result could aid drug discovery efforts aimed at inhibiting blood vessel growth in tumors (Nature doi:10.1038/ nature22322, 2017). IntheUS,duringtheObamaadministration, industry and academia joined forces to apply AI to accelerate drug discovery as part of the CancerMoonshotinitiative(Nat.Biotechnol.34, 119, 2016). The Accelerating Therapeutics for Opportunities in Medicine (ATOM), launched in January 2016, marries computational and experimental approaches, with Brentford, UK-based GlaxoSmithKline, participating with Lawrence Livermore National Laboratory in Livermore, California, and the US National Cancer Institute. The computational portion of the process, which includes deep-learning and other AI algorithms, will be tested in the first two years. In the third year, “we hope to start on day one with a disease hypothesis and on day 365 to deliver a drug candidate,” says MarthaHead,GlaxoSmithKline’shead,insights from data. Table 1 Selected collaborations in the AI-drug discovery space AI company/ location Technology Announced partner/ location Indication(s) Deal date Atomwise Deep-learning screening from molecular structure data Merck Malaria 2015 BenevolentAI Deep-learning and natural language processing of research literature Janssen Pharmaceutica (Johnson & Johnson), Beerse, Belgium Multiple November 8, 2016 Berg, Framingham, Massachusetts Deep-learning screening of biomarkers from patient data None Multiple N/A Exscientia Bispecific compounds via Bayesian models of ligand activity from drug discovery data Sanofi Metabolic diseases May 9, 2017 GNS Healthcare Bayesian probabilistic inference for investigating efficacy Genentech Oncology June 19, 2017 Insilico Medicine Deep-learning screening from drug and disease databases None Age-related diseases N/A Numerate Deep learning from pheno- typic data Takeda Oncology, gastro- enterology and central nervous system disorders June 12, 2017 Recursion, Salt Lake City, Utah Cellular phenotyping via image analysis Sanofi Rare genetic diseases April 25, 2016 twoXAR, Palo Alto, California Deep-learning screening from literature and assay data Santen Pharmaceuticals, Osaka, Japan Glaucoma February 23, 2017 N/A, none announced. Source: companies’ websites. N E W S
- 229.
- 230. •Digiceutical = digital+ pharmaceutical •"chemical 과 protein에 이어서 digital drug 이 세번째 종류의 신약이 될 것이다” •digital drug 은 크게 두 가지 종류 •기존의 약을 아예 대체 •기존 약을 강화(augment)
- 231. PTSD (외상 후스트레스 장애)
- 232.
- 234. scores at baseline,post treatment and 3-month follow-up are in Fig group, mean Beck Anxiety Inventory scores significantly decrea (9.5) to 11.9 (13.6), (t=3.37, df=19, p < .003) and mean PHQ-9 decreased 49% from 13.3 (5.4) to 7.1 (6.7), (t=3.68, df=19, p < 0.00 Figure 4. PTSD Checklist scores across treatment Figure 5. BAI and PH The average number of sessions for this sample was just under successful treatment completers had documented mild and mode injuries, which suggest that this form of exposure can be useful PTSD Checklist scores across treatment • 연구 결과 20명의 환자들은 전반적으로 유의미한 개선을 보임 • 환자들 전체의 PCL-M 수치가 평균 54.4에서 35.6으로 감소 • 20명 중 16명은 치료 직후에 더 이상 PTSD 를 가지지 않은 것으로 나타남 • 치료가 끝난지 3개월 후에 환자들의 상태는 유지 http://www.ncbi.nlm.nih.gov/pubmed/19377167
- 235. reatment and 3-monthfollow-up are in Figure 4. For this same iety Inventory scores significantly decreased 33% from 18.6 =3.37, df=19, p < .003) and mean PHQ-9 (depression) scores 3 (5.4) to 7.1 (6.7), (t=3.68, df=19, p < 0.002) (see Figure 5). ores across treatment Figure 5. BAI and PHQ-Depression scores r of sessions for this sample was just under 11. Also, two of the mpleters had documented mild and moderate traumatic brain that this form of exposure can be usefully applied with this BAI and PHQ-Depression scores • 벡 불안 지수는 평균 18.6에서 11.9로 33% 감소 • PHQ-9 우울증 지수 역시 13.3에서 7.1로 49% 감소 • 경미한 외상성 뇌손상 (traumatic brain injury) 환자 2명에도 유의미한 효과 http://www.ncbi.nlm.nih.gov/pubmed/19377167
- 237. RespeRate •FDA 승인 받은유일한 비약물 고혈압 치료법 •sessions of therapeutic breathing 을 통해서 혈압 강하 효과 •15분씩 일주일에 a few times 활용하면 significant blood pressure reduction 증명 •전세계 25만 명 이상 사용
- 238.
- 239. 2breathe •디지털 기기 중,수면 유도 목적으로는 2breathe가 유일 •고혈압 치료기기의 ‘부작용’으로 수면 유도 효과 발견 •안전성은 수십만 명의 환자에게 임상 시험 통해서 증명 •교감신경의 활성화를 줄임으로써 사용자의 릴렉스와 수면을 유도
- 241.
- 243. Effects of virtualreality-based rehabilitation on distal upper extremity function and health-related quality of life: a single-blinded, randomized controlled trial ments at T2 and 23 completed the follow-up assessments at T3. During the study, 5 and 8 participants from the SG and CON groups, respectively, did not complete the inter- vention programs. The sample sizes at the assessment time points are presented in Fig. 2. There were no serious ad- verse events, and only 1 participant from the CON group dropped out owing to dizziness, which was unrelated to the intervention. Thus, most of the study withdrawals were related to uncooperativeness, and the number was higher than that hypothesized in the study design. At baseline, dist: F = 4.64, df = 1.38, P = 0.024). Secondary outcomes Jebsen–Taylor hand function test The JTT scores of the SG and CON groups are presented in Table 2. There were no significant differences in the JTT-total, JTT-gross, and JTT-fine scores between the 2 groups at T0. The post-hoc test found that there were sig- nificant improvements in the JTT-total, JTT-gross, and JTT-fine scores in the SG group during the intervention Fig. 2 Flowchart of the participants through the study. Abbreviations: SG, Smart Glove; CON, conventional intervention Shin et al. Journal of NeuroEngineering and Rehabilitation (2016) 13:17 Shin et al. Journal of NeuroEngineering and Rehabilitation (2016) 13:17
- 244. Effects of virtualreality-based rehabilitation on distal upper extremity function and health-related quality of life: a single-blinded, randomized controlled trial composite SIS score (F = 5.76, df = 1.0, P = 0.021) and the overall SIS score (F = 6.408, df = 1.0, P = 0.015). Moreover, among individual domain scores, the Time × standard OT than using amount-matched conventional re- habilitation, without any adverse events, in stroke survivors. Additionally, this study noted improvements in the SIS- Fig. 3 Mean and standard errors for the FM scores in the SG and CON groups. Abbreviations: FM, Fugl–Meyer assessment, SG, Smart Glove; CON, conventional intervention Fig. 4 Mean and standard errors for the JTT scores in the SG and CON groups. Abbreviations: JTT, Jebsen–Taylor hand function test; SG, Smart Glove; CON, conventional intervention Shin et al. Journal of NeuroEngineering and Rehabilitation (2016) 13:17 Page 7 of 10 composite SIS score (F = 5.76, df = 1.0, P = 0.021) and the overall SIS score (F = 6.408, df = 1.0, P = 0.015). standard OT than using amount-matched conventional re- habilitation, without any adverse events, in stroke survivors. Fig. 3 Mean and standard errors for the FM scores in the SG and CON groups. Abbreviations: FM, Fugl–Meyer assessment, SG, Smart Glove; CON, conventional intervention Fig. 4 Mean and standard errors for the JTT scores in the SG and CON groups. Abbreviations: JTT, Jebsen–Taylor hand function test; SG, Smart Glove; CON, conventional intervention Shin et al. Journal of NeuroEngineering and Rehabilitation (2016) 13:17 Page 7 of 10 Shin et al. Journal of NeuroEngineering and Rehabilitation (2016) 13:17
- 247. Weight loss efficacyof a novel mobile Diabetes Prevention Program delivery platform with human coaching Andreas Michaelides, Christine Raby, Meghan Wood, Kit Farr, Tatiana Toro-Ramos To cite: Michaelides A, Raby C, Wood M, et al. Weight loss efficacy of a novel mobile Diabetes Prevention Program delivery platform with human coaching. BMJ Open Diabetes Research and Care 2016;4:e000264. doi:10.1136/bmjdrc-2016- 000264 Received 4 May 2016 Revised 19 July 2016 Accepted 11 August 2016 Noom, Inc., New York, New York, USA Correspondence to Dr Andreas Michaelides; andreas@noom.com ABSTRACT Objective: To evaluate the weight loss efficacy of a novel mobile platform delivering the Diabetes Prevention Program. Research Design and Methods: 43 overweight or obese adult participants with a diagnosis of prediabetes signed-up to receive a 24-week virtual Diabetes Prevention Program with human coaching, through a mobile platform. Weight loss and engagement were the main outcomes, evaluated by repeated measures analysis of variance, backward regression, and mediation regression. Results: Weight loss at 16 and 24 weeks was significant, with 56% of starters and 64% of completers losing over 5% body weight. Mean weight loss at 24 weeks was 6.58% in starters and 7.5% in completers. Participants were highly engaged, with 84% of the sample completing 9 lessons or more. In-app actions related to self-monitoring significantly predicted weight loss. Conclusions: Our findings support the effectiveness of a uniquely mobile prediabetes intervention, producing weight loss comparable to studies with high engagement, with potential for scalable population health management. INTRODUCTION Lifestyle interventions,1 including the National Diabetes Prevention Program (NDPP) have proven effective in preventing type 2 diabetes.2 3 Online delivery of an adapted NDPP has resulted in high levels of engagement, weight loss, and improvements in glycated hemoglobin (HbA1c).4 5 Prechronic and chronic care efforts delivered by other means (text and emails,6 nurse support,7 DVDs,8 community care9 ) have also been successful in promoting behavior change, weight loss, and glycemic control. One study10 adapted the NDPP to deliver the first part of the curriculum in-person and the remaining sessions through a mobile app, and found 6.8% weight loss at 5 months. Mobile health poses a promising means of delivering prechronic and chronic care,11 12 and provides a scalable, convenient, and accessible method to deliver the NDPP. The weight loss efficacy of a completely mobile delivery of a structured NDPP has not been tested. The main aim of this pilot study was to evaluate the weight loss efficacy of Noom’s smartphone-based NDPP-based cur- ricula with human coaching in a group of overweight and obese hyperglycemic adults receiving 16 weeks of core, plus postcore cur- riculum. In this study, it was hypothesized that the mobile DPP could produce trans- formative weight loss over time. RESEARCH DESIGN AND METHODS A large Northeast-based insurance company offered its employees free access to Noom Health, a mobile-based application that deli- vers structured curricula with human coaches. An email or regular mail invitation with information describing the study was sent to potential participants based on an elevated HbA1c status found in their medical records, reflecting a diagnosis of prediabetes. Interested participants were assigned to a virtual Centers for Disease Control and Prevention (CDC)-recognized NDPP master’s level coach. Key messages ▪ To the best of our knowledge, this study is the first fully mobile translation of the Diabetes Prevention Program. ▪ A National Diabetes Prevention Program (NDPP) intervention delivered entirely through a smart- phone platform showed high engagement and 6-month transformative weight loss, comparable to the original NDPP and comparable to trad- itional in-person programmes. ▪ This pilot shows that a novel mobile NDPP inter- vention has the potential for scalability, and can address the major barriers facing the widespread translation of the NDPP into the community setting, such as a high fixed overhead, fixed locations, and lower levels of engagement and weight loss. BMJ Open Diabetes Research and Care 2016;4:e000264. doi:10.1136/bmjdrc-2016-000264 1 Open Access Research group.bmj.comon April 27, 2017 - Published byhttp://drc.bmj.com/Downloaded from •Noom Coach 앱이 체중 감량을 위해서 효과적임을 증명 •완전히 모바일로 이뤄진 최초의 당뇨병 예방 연구 •43명의 전당뇨단계에 있는 과체중이나 비만 환자를 대상 •24주간 Noom Coach의 앱과 모바일 코칭을 제공 •그 결과 64% 의 참가자들이 5-7% 의 체중 감량 효과 •84%에 달하는 사람들이 마지막까지 이 6개월 간의 프로그램에 참여
- 248. www.nature.com/scientificreports Successful weight reduction andmaintenance by using a smartphone application in those with overweight and obesity SangOukChin1,* ,Changwon Keum2,* , JunghoonWoo3 , Jehwan Park2 , Hyung JinChoi4 , Jeong-taekWoo5 & SangYoul Rhee5 A discrepancy exists with regard to the effect of smartphone applications (apps) on weight reduction due to the several limitations of previous studies.This is a retrospective cohort study, aimed to investigate the effectiveness of a smartphone app on weight reduction in obese or overweight individuals, based on the complete enumeration study that utilized the clinical and logging data entered by NoomCoach app users betweenOctober 2012 andApril 2014.A total of 35,921 participants were included in the analysis, of whom 77.9% reported a decrease in body weight while they were using the app (median 267 days; interquartile range=182). Dinner input frequency was the most important factor for successful weight loss (OR=10.69; 95%CI=6.20–19.53; p<0.001), and more frequent input of weight significantly decreased the possibility of experiencing the yo-yo effect (OR=0.59, 95%CI=0.39–0.89; p<0.001).This study demonstrated the clinical utility of an app for successful weight reduction in the majority of the app users; the effects were more significant for individuals who monitored their weight and diet more frequently. Obesity is a global epidemic with a rapidly increasing prevalence worldwide1,2 . As obese individuals experience significantly higher mortality when compared with the non-obese population3,4 , this phenomenon poses a sig- nificant socioeconomic burden, necessitating strategies to manage overweight and prevent obesity5 . Although numerous interventions such as life style modification including exercise6–10 , and pharmacotherapy11–13 have been shown effective for both the prevention and treatment of obesity, some of these methods were found to have a limitation which required substantial financial inputs and repeated time-consuming processes14,15 . Recently, as the number of smartphone users is increasing dramatically, many investigators have attempted to implement smartphone applications (app) for health promotion16–19 . Consequently, many smartphone apps have demonstrated at least partial efficacy in promoting successful weight reduction according to the number of previous studies20–24 . However, due to the limitations associated with study design such as small-scale studies and short investigation periods, a discrepancy exists with regard to the effect of apps on weight reduction20,21,23 . Even systemic reviews which investigated the efficacy of mobile apps for weight reduction reported more or less inconsistent results; Flores Mateo et al. reported a significant weight loss by mobile phone app intervention when compared with control groups25 whereas Semper et al. reported that four of the six studies included in the analysis showed no significant difference of weight reduction between comparison groups26 . Thus, the aim of this study was to investigate the effectiveness of a smartphone app on weight reduction in obese or overweight individuals Recei e : 0 pri 016 Accepte : 15 eptem er 016 Pu is e : 0 o em er 016 OPEN •스마트폰 앱이 체중 감량에 도움을 줄 수 있는가? •2012년부터 2014년 까지 최소 6개월 이상 애플리케이션을 사용 •80여 국가(미국, 독일, 한국, 영국, 일본 등)에서 모집된 35,921명의 데이터 •애플리케이션 평균 사용기간은 267일 Chin et al. Sci Rep 2016
- 249. www.nature.com/scientificreports/ Figure 1. Distributionof weight loss among app users. Percentages (and 95% CIs) of participants achieving <5%, 5–10%, 10–15%, 15–20% and >20% weight loss relative to baseline at the end of the 6-month trial period. Data are reported as the mean±SD. Univariate Linear Regression p-value Multivariate Linear Regression p-valueβ (95% CI) β (95% CI) Gender (male) 0.60 (0.54, 0.66) <0.001 0.71 (0.65, 0.77) <0.001 Age 0.01 (0.008, 0.013) <0.001 −0.026 (−0.03, −0.02) <0.001 Follow-up Days −0.001 (−0.001, −0.001) <0.001 0.00 (0.00, 0.00) 0.886 Baseline BMI 0.146 (0.143, 0.150) <0.001 0.165 (0.161, 0.168) <0.001 Successful weight reduction and maintenance by using a smartphone application in those with overweight and obesity Chin et al. Sci Rep 2016 •대상자의 약 77.9%에서 성공적인 체중감량 효과를 확인 •이 중 23%는 본인 체중의 10% 이상 감량에 성공 •앱의 사용이 약물 치료 등 다른 비만 관리 기법에 비해 체중 감량 효과가 뒤쳐지지 않음
- 250. Successful weight reduction and maintenance by using a smartphone application in those with overweight and obesity Chin et al.Sci Rep 2016 •체중을 자주 기록하고 저녁식사를 자주 입력한 사용자의 체중감량 효과가 가장 높았음 •비만 관리에서 강조되던 생활 습관의 중요성을 글로벌 스케일의 데이터로 증명 nature.com/scientificreports/ Diabetes Prevention Program (DPP)-intensive lifestyle intervention is one such method, designed to produce clinically significant weight reduction in adults with prediabetes, proving its effectiveness for both weight loss and cardiometabolic outcomes8 . In addition, life style modification has been shown to be effective for reducing body weight and cardiovascular risk6–10 ; however, each of these studies had important limitations, particularly in that some of them were resource intensive, expensive, and time-consuming14,15 . Frequent group and individual Univariate Logistic Regression Wald Test p-value Multivariate Logistic Regression Wald Test p-valueOR (95% CI) OR (95% CI) Gender (male) 1.44 (1.29, 1.60) <0.001 2.05 (1.79, 2.36) <0.001 Age 0.99 (0.99, 1.00) 0.002 0.97 (0.95, 0.97) <0.001 Follow-up Days 1.00 (1.000, 1.00) 0.627 — — Baseline BMI 1.10 (1.09, 1.11) <0.001 1.13 (1.12, 1.14) <0.001 Weight input frequency (n/person-day) 2.85 (2.20, 3.70) <0.001 3.0 (2.21, 4.08) <0.001 Breakfast input frequency (n/person-day) 3.15 (2.72, 3.66) <0.001 0.36 (0.22, 0.57) <0.001 Lunch input frequency (n/person-day) 3.98 (3.42, 4.64) <0.001 1.14 (0.57, 2.28) 0.718 Dinner input frequency (n/person-day) 4.86 (4.16, 5.68) <0.001 10.69 (6.20, 18.53) <0.001 Breakfast calories (kcal/person-day) 1.00 (1.00, 1.00) <0.001 1.00 (1.00, 1.00) <0.001 Lunch calories (kcal/person-day) 1.00 (1.00, 1.00) <0.001 1.00 (1.00, 1.00) <0.001 Dinner calories (kcal/person-day) 1.00 (1.00, 1.00) 0.105 1.00 (1.00, 1.00) <0.001 Exercise input frequency (n/person-day) 4.02 (3.30, 4.90) <0.001 2.49 (1.96, 3.17) <0.001 Exercise calories expenditure (kcal/person-day) 1.00 (1.00, 1.00) <0.001 1.00 (1.00, 1.00) 0.085 Table 4. Factors contributing to being a success or a partial success against stationary subgroup. Abbreviations: BMI, body mass index; OR, odds ratio; CI, confidence interval.
- 251. •미국 CDC의 당뇨병예방 프로그램(DPP)으로 공식 인증 •CDC에서 fully recognised 된 첫번째 ‘virtual provider’ •2018년 1월부터 CMS(Centers for Medicare&Medicaid Services)의 보험 수가를 적용 •메디케어 1인당 2년에 성취도에 따라 $630 까지 지급 •B2B 사업으로도 확대 예정 "눔은 OEM(주문자상표부착생산) 업체로서 라이선스를 사간 기업에 모바일 플랫폼과 건강관리 코치들, 교육프로그램 등을 종합적으로 제공한다"
- 252.
- 253. 우울증 치료 임상결과 1 임상 기간 : 2014년 10월 ~ 2016년 12월 N=96, 1회 30분 자극 Severe Moderate Mild 10 20 30 40 Beck Depression Inventory II 6주 42회 연속 복용SSRI Ybrain 5회 1회 1회5회 0 10 20 30 40 BASELINE 2 WEEK 4 WEEK 6 WEEK MADRS 6주 42회 연속 복용 Ybrain SSRI 5회 1회 1회5회 Severe Moderate Mild None Primary Outcome: 몽고메리-아스퍼그 우울평가척도(MADRS) Secondary Outcome: Beck 우울 척도(Beck Depression Inventory II) Courtesy of 이기원 대표님, YBrain •국내 96명 환자를 대상으로 2년간 double-blinded randomised 임상 연구 실시 •실험군: 가짜 약+ 진짜 자극기기 •대조군: 진짜 약 + 가짜 자극기기 •Primary Outcome인 MADRS 스케일에서 기기가 약에 조금 못 미치는 결과
- 254. 우울증 치료 임상결과 1 임상 기간 : 2014년 10월 ~ 2016년 12월 N=96, 1회 30분 자극 Severe Moderate Mild 10 20 30 40 Beck Depression Inventory II 6주 42회 연속 복용SSRI Ybrain 5회 1회 1회5회 0 10 20 30 40 BASELINE 2 WEEK 4 WEEK 6 WEEK MADRS 6주 42회 연속 복용 Ybrain SSRI 5회 1회 1회5회 Severe Moderate Mild None Primary Outcome: 몽고메리-아스퍼그 우울평가척도(MADRS) Secondary Outcome: Beck 우울 척도(Beck Depression Inventory II) Courtesy of 이기원 대표님, YBrain •Primary Outcome인 MADRS에서 기존 약물에 비해서 약간 효능이 적게 나옴 •Secondary Outcome인 BDI 에 대해서는 기존 약물과 동등하게 나옴 •이러한 결과에 따라서 식약처에서 ‘3등급 보조의료기기’ 로 인허가 •따라서, 원칙적으로는 기존에 우울증 약을 복용하는 환자를 대상으로 사용하게 될 것임
- 255. •경두개 직류자극치료술(tDCS) •2017년 3월국내 최초로 식약처의 3등급 보조의료기기 허가 •7월에는 유럽 CE허가를 받을 예정 •2~3년 내 FDA 허가를 받는 것을 목표 •추가 임상 연구 예정 •우울증 •독거 노인 우울증 치료 시범 사업 진행 중 •10월부터 하버드 의대와 아시아 지역 500명 대상의 임상 예정 •경도인지장애 임상 예정 •조현병 1차 임상 마무리 + 논문 출판 예정 •신의료기술평가 진행 예정
- 256.
- 257. • reSET® wasevaluated in a clinical trial of 507 patients with SUD across 10 treatment centers nation-wide over 12 weeks.* • In patients who were dependent on stimulants, marijuana, cocaine, or alcohol (n=395), 58.1% of patients receiving reSET®* were abstinent in study weeks 9-12, while 29.8% of patients receiving face-to-face therapy alone were abstinent during the same time frame (p<0.01). • Participants who tested positive for drug use at the start of the study (n=191), 26.7% of patients receiving reSET®* were abstinent in study weeks 9-12, while 3.2% of patients receiving traditional face-to-face therapy were abstinent during the same time frame (p<0.01). Pear Therapeutics Campbell et al. Am J Psychiatry. 2014.
- 258. Campbell et al.Am J Psychiatry. 2014. Pear Therapeutics • Patients receiving reSET® showed statistically significant improvement in retention compared to face-to-face therapy alone (p=0.0316).At the end of 12 weeks of treatment 59% of patients receiving face-to-face therapy were retained in the study compared to 67% of patients receiving reSET®.
- 259. Pear Therapeutics •최초로 스마트폰앱이 digital therapeutics 로 질병 치료 목적으로 FDA de novo clearance (기기 없이 '앱'만으로 구성된 시스템이 '질병 치료' 목적으로 허가 받은 것은 최초) •Pear Therapeutics의 reSET 이라는 시스템으로 각종 중독을 치료하는 목적의 앱 •12주에 걸쳐서 대마, 코카인, 알콜 중독에 대한 중독과 의존성을 치료
- 260. Three Steps toImplement Digital Medicine • Step 1. Measure the Data • Step 2. Collect the Data • Step 3. Insight from the Data digital therapeutics +
- 262. Feedback/Questions • E-mail: yoonsup.choi@gmail.com •Blog: http://www.yoonsupchoi.com • Facebook: 최윤섭 디지털 헬스케어 연구소