SlideShare a Scribd company logo
1 of 35
Download to read offline
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Ujjwal Ratan
Principal AI/ML SA, GLS
AWS
L F S 3 0 3
Mengqi Jin
Machine Learning Scientist
AWS
Bryan Coon
Sr. Manager, Solutions
Architecture, GLS
AWS
Natural Language Processing and Mortality Risk Prediction Using
Amazon Comprehend Medical & Amazon SageMaker
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Agenda
What is Amazon Comprehend Medical?
What is Amazon SageMaker?
Workshop use case
Workshop architecture
Hands on
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Amazon Comprehend Medical
A new HIPAA-eligible service that uses machine learning to extract medical information with high
accuracy, reducing the cost, time and effort of processing large amounts of unstructured medical text
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Amazon Comprehend Medical
Entities
• Medication
• Medical condition
• Test, Treatments and Procedures
• Anatomy
• Protected Health Information (PHI)
Relationship Extraction
• Medication and dosage
• Test and result
• Many more
Entity Traits
• Negation
• Diagnosis, Sign or
Symptom
Protected Health Information Identification
(PHId API)
Distill a complex process into a simple API call
Medical Named Entity and
Relationship Extraction (NERe API)
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Protected Health Information (PHI)
Mr. Doe : Name
63 : Age
Comprehend Medical - NERe API
Mr. Doe is a 63-year-old gentleman with
coronary artery disease and hypertension.
CURRENT MEDICATIONS: taking a dose of
LIPITOR 20 mg once daily.
aws comprehend-medical detect-entities --region
us-east-1 --text “<Insert Text Here>”
Named Entities
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Comprehend Medical – NERe API
CURRENT MEDICATIONS: taking a dose of
LIPITOR 20 mg once daily
aws comprehend-medical detect-entities --region
us-east-1 --text “<Insert Text Here>”
Relationship Extraction
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Amazon Comprehend – NERe API
What are Entity Traits?
(1) Negation
(2) Diagnosis, sign or symptom
Rash
Discontinue Abraxane, patient denies
taking Tylenol 325 mg and is not taking
calcium carbonate, patient also stopped
taking cholecalciferol 1,000 units PO
Entity Traits
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Amazon Comprehend – PHId API
Mr. Doe is a 63-year-old gentleman with
coronary artery disease and hypertension. He
currently lives in Anytown and works as a
teacher. He currently lives in Anytown and
works as a teacher. His PCP, Dr. Stiles, works at
the University of Anystate.
aws comprehend-medical detect-phi --region us-
east-1 --text “<Insert Text Here>”
In addition to extracting PHI, the
PHId API identifies relevant patient
identifiers described in HIPAA Safe
Harbor method of de-identification
Personal Health Information
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
“For cancer patients and the researchers dedicated to curing them,
time is the limiting resource. The process of developing clinical
trials and connecting them with the right patients requires
research teams to sift through and label mountains of clinical
record data. The Amazon Comprehend Medical service reduces
this time burden from hours to seconds. This is a vital step toward
getting researchers rapid access to the information they need
when they need it so they can advance lifesaving therapies for
patients.”
Matthew Trunnell,
Chief Information Officer,
Fred Hutchinson Cancer Research Center
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
A fully managed service that enables data scientists and developers to quickly and easily
build machine-learning based models into production smart applications.
Amazon SageMaker
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
End-to-End
Machine Learning
Platform
Zero setup Flexible Model
Training
Pay by the second
Amazon SageMaker
Build, train, and deploy machine learning models at scale
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
One-click training
for ML, DL, and
custom algorithms
Easier training with
hyperparameter
optimization
Highly optimized
machine learning
algorithms
Deployment
without
engineering effort
Fully managed
hosting at scale
BuildPre-built
notebook
instances
Deploy
Train
Amazon SageMaker
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Business problem
Predict an ICU patient’s risk of mortality
Why is this important?
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
This workshop has four sections
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Go to the following link to begin setup
https://bit.ly/2KcWj6R
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Architecture diagram
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Mortality risk prediction
Amazon
ECR
Amazon
SageMaker
Amazon
S3
Data
Algorithm
Training Inference
Model Artifacts Prediction Outputs
Amazon
S3
Doc2Vec
Note Embeddings
1 32
Notebooks
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Mortality risk prediction
0 hr 12 hr 24 hr 36 hr 48 hr … … … … Discharge
Observation
Window
Prediction
Window
Death? DischargeICU Admission Prediction
• Goal: Early detection of at-risk
patients in ICU
• Using first 48 hrs data of
admission to predict in-hospital
mortality
• Numerical Variables:
• Vital Signs
• Labs
• Bedside Measurements
• Unstructured Notes:
• Nursing Notes
• Physician Notes
• ECG Notes
• Other 12 types of notes
Mortality is one of the metrics
used by CMS to evaluate quality
of care and hospital ranking
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Risk modeling with text information
Multimodal
Embedding
Raw Notes
Model
Artifacts
Doc2VecC
Medication
Diagnosis
Treatment/
Test/
Procedure
Recurrent
Neural Network
Amazon
Comprehend
Medical
Extracted Medical Entities
Concatenated
entities
Step 1: Text Processing
Step 2: Training Step 3: Offline Testing
Evaluation
Metrics
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Risk modeling with text information
Step 1: Generate Document Embeddings
• Use Doc2VecC to generate embeddings for 4000 medical notes
• Save trained embedding dictionary on Amazon Simple Storage Service (Amazon S3)
Input
Raw notes saved in csv files
Output
NumPy dictionary of document embeddings keyed by Note ID
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Risk modeling with text information
Step 2: Amazon SageMaker Training Job
• Run the algorithm over the training data
• Save the output model and select the best one to save it on Amazon S3
Input
Time series data
Pre-trained document embeddings
Output
Model artifacts saved in model.tar.gz file
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Risk modeling with text information
Step 3: Amazon SageMaker BatchTransform Job
• Making inference on test data set
• Evaluate model performance
Input
Selected best model saved in model.tar.gz file
Time series data
Pre-trained document embeddings
Output
Prediction for each time series record
Evaluation metrics
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Related breakouts
Thursday November 29th
Introducing Amazon Comprehend Medical
12:15 – 1:15 PM | Aria West Level 3 Juniper 4
Thank you!
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Ujjwal Ratan
ujjwalr@amazon.com
Bryan Coon
bccn@amazon.com
Mengqi Jin
maggjin@amazon.com
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.

More Related Content

What's hot

Manufacturing Better Smiles with Neural Networks (HLC311) - AWS re:Invent 2018
Manufacturing Better Smiles with Neural Networks (HLC311) - AWS re:Invent 2018Manufacturing Better Smiles with Neural Networks (HLC311) - AWS re:Invent 2018
Manufacturing Better Smiles with Neural Networks (HLC311) - AWS re:Invent 2018Amazon Web Services
 
Build, Train, and Deploy ML Models with Amazon SageMaker (AIM410-R2) - AWS re...
Build, Train, and Deploy ML Models with Amazon SageMaker (AIM410-R2) - AWS re...Build, Train, and Deploy ML Models with Amazon SageMaker (AIM410-R2) - AWS re...
Build, Train, and Deploy ML Models with Amazon SageMaker (AIM410-R2) - AWS re...Amazon Web Services
 
Enabling Patient Centricity for Pfizer through AWS Cloud (LFS301-S-i) - AWS r...
Enabling Patient Centricity for Pfizer through AWS Cloud (LFS301-S-i) - AWS r...Enabling Patient Centricity for Pfizer through AWS Cloud (LFS301-S-i) - AWS r...
Enabling Patient Centricity for Pfizer through AWS Cloud (LFS301-S-i) - AWS r...Amazon Web Services
 
Building IoT Devices for Regulated Industries (LFS304-i) - AWS re:Invent 2018
Building IoT Devices for Regulated Industries (LFS304-i) - AWS re:Invent 2018Building IoT Devices for Regulated Industries (LFS304-i) - AWS re:Invent 2018
Building IoT Devices for Regulated Industries (LFS304-i) - AWS re:Invent 2018Amazon Web Services
 
DevSecOps: Instituting Cultural Transformation for Public Sector Organization...
DevSecOps: Instituting Cultural Transformation for Public Sector Organization...DevSecOps: Instituting Cultural Transformation for Public Sector Organization...
DevSecOps: Instituting Cultural Transformation for Public Sector Organization...Amazon Web Services
 
Architecting for GxP Compliance in Life Sciences (LFS316) - AWS re:Invent 2018
Architecting for GxP Compliance in Life Sciences (LFS316) - AWS re:Invent 2018Architecting for GxP Compliance in Life Sciences (LFS316) - AWS re:Invent 2018
Architecting for GxP Compliance in Life Sciences (LFS316) - AWS re:Invent 2018Amazon Web Services
 
Detecting Financial Market Manipulation Using Machine Learning (AIM347) - AWS...
Detecting Financial Market Manipulation Using Machine Learning (AIM347) - AWS...Detecting Financial Market Manipulation Using Machine Learning (AIM347) - AWS...
Detecting Financial Market Manipulation Using Machine Learning (AIM347) - AWS...Amazon Web Services
 
Can Deep Learning Be Used to Treat Disease and Measure Product Value? (LFS310...
Can Deep Learning Be Used to Treat Disease and Measure Product Value? (LFS310...Can Deep Learning Be Used to Treat Disease and Measure Product Value? (LFS310...
Can Deep Learning Be Used to Treat Disease and Measure Product Value? (LFS310...Amazon Web Services
 
Build Deep Learning Applications Using PyTorch and Amazon SageMaker (AIM432-R...
Build Deep Learning Applications Using PyTorch and Amazon SageMaker (AIM432-R...Build Deep Learning Applications Using PyTorch and Amazon SageMaker (AIM432-R...
Build Deep Learning Applications Using PyTorch and Amazon SageMaker (AIM432-R...Amazon Web Services
 
Hollywood's Cloud-Based Content Lakes: Modernized Media Archives (MAE203) - A...
Hollywood's Cloud-Based Content Lakes: Modernized Media Archives (MAE203) - A...Hollywood's Cloud-Based Content Lakes: Modernized Media Archives (MAE203) - A...
Hollywood's Cloud-Based Content Lakes: Modernized Media Archives (MAE203) - A...Amazon Web Services
 
[NEW LAUNCH!] Introducing Amazon Forecast (AIM344) - AWS re:Invent 2018
[NEW LAUNCH!] Introducing Amazon Forecast  (AIM344) - AWS re:Invent 2018[NEW LAUNCH!] Introducing Amazon Forecast  (AIM344) - AWS re:Invent 2018
[NEW LAUNCH!] Introducing Amazon Forecast (AIM344) - AWS re:Invent 2018Amazon Web Services
 
Build a "Who's Who" App for Your Media Content (AIM409) - AWS re:Invent 2018
Build a "Who's Who" App for Your Media Content (AIM409) - AWS re:Invent 2018Build a "Who's Who" App for Your Media Content (AIM409) - AWS re:Invent 2018
Build a "Who's Who" App for Your Media Content (AIM409) - AWS re:Invent 2018Amazon Web Services
 
透過最新的 AWS 服務在 2019 年為您的業務轉型 (Level 200)
透過最新的 AWS 服務在 2019 年為您的業務轉型 (Level 200)透過最新的 AWS 服務在 2019 年為您的業務轉型 (Level 200)
透過最新的 AWS 服務在 2019 年為您的業務轉型 (Level 200)Amazon Web Services
 
Sequence-to-Sequence Modeling with Apache MXNet, Sockeye, and Amazon SageMake...
Sequence-to-Sequence Modeling with Apache MXNet, Sockeye, and Amazon SageMake...Sequence-to-Sequence Modeling with Apache MXNet, Sockeye, and Amazon SageMake...
Sequence-to-Sequence Modeling with Apache MXNet, Sockeye, and Amazon SageMake...Amazon Web Services
 
Build an Intelligent Multi-Modal User Agent with Voice and NLU (AIM340) - AWS...
Build an Intelligent Multi-Modal User Agent with Voice and NLU (AIM340) - AWS...Build an Intelligent Multi-Modal User Agent with Voice and NLU (AIM340) - AWS...
Build an Intelligent Multi-Modal User Agent with Voice and NLU (AIM340) - AWS...Amazon Web Services
 
Build an ETL Pipeline to Analyze Customer Data (AIM416) - AWS re:Invent 2018
Build an ETL Pipeline to Analyze Customer Data (AIM416) - AWS re:Invent 2018Build an ETL Pipeline to Analyze Customer Data (AIM416) - AWS re:Invent 2018
Build an ETL Pipeline to Analyze Customer Data (AIM416) - AWS re:Invent 2018Amazon Web Services
 
Deep Dive on Amazon S3 Storage Classes: Creating Cost Efficiencies across You...
Deep Dive on Amazon S3 Storage Classes: Creating Cost Efficiencies across You...Deep Dive on Amazon S3 Storage Classes: Creating Cost Efficiencies across You...
Deep Dive on Amazon S3 Storage Classes: Creating Cost Efficiencies across You...Amazon Web Services
 
AI/ML with Data Lakes: Counterintuitive Consumer Insights in Retail (RET206) ...
AI/ML with Data Lakes: Counterintuitive Consumer Insights in Retail (RET206) ...AI/ML with Data Lakes: Counterintuitive Consumer Insights in Retail (RET206) ...
AI/ML with Data Lakes: Counterintuitive Consumer Insights in Retail (RET206) ...Amazon Web Services
 
Go Fast and Remain Secure: How Millennium Enables Developers and Upholds Secu...
Go Fast and Remain Secure: How Millennium Enables Developers and Upholds Secu...Go Fast and Remain Secure: How Millennium Enables Developers and Upholds Secu...
Go Fast and Remain Secure: How Millennium Enables Developers and Upholds Secu...Amazon Web Services
 
Leadership Session: Innovation-Driven Healthcare (HLC201-iL) - AWS re:Invent ...
Leadership Session: Innovation-Driven Healthcare (HLC201-iL) - AWS re:Invent ...Leadership Session: Innovation-Driven Healthcare (HLC201-iL) - AWS re:Invent ...
Leadership Session: Innovation-Driven Healthcare (HLC201-iL) - AWS re:Invent ...Amazon Web Services
 

What's hot (20)

Manufacturing Better Smiles with Neural Networks (HLC311) - AWS re:Invent 2018
Manufacturing Better Smiles with Neural Networks (HLC311) - AWS re:Invent 2018Manufacturing Better Smiles with Neural Networks (HLC311) - AWS re:Invent 2018
Manufacturing Better Smiles with Neural Networks (HLC311) - AWS re:Invent 2018
 
Build, Train, and Deploy ML Models with Amazon SageMaker (AIM410-R2) - AWS re...
Build, Train, and Deploy ML Models with Amazon SageMaker (AIM410-R2) - AWS re...Build, Train, and Deploy ML Models with Amazon SageMaker (AIM410-R2) - AWS re...
Build, Train, and Deploy ML Models with Amazon SageMaker (AIM410-R2) - AWS re...
 
Enabling Patient Centricity for Pfizer through AWS Cloud (LFS301-S-i) - AWS r...
Enabling Patient Centricity for Pfizer through AWS Cloud (LFS301-S-i) - AWS r...Enabling Patient Centricity for Pfizer through AWS Cloud (LFS301-S-i) - AWS r...
Enabling Patient Centricity for Pfizer through AWS Cloud (LFS301-S-i) - AWS r...
 
Building IoT Devices for Regulated Industries (LFS304-i) - AWS re:Invent 2018
Building IoT Devices for Regulated Industries (LFS304-i) - AWS re:Invent 2018Building IoT Devices for Regulated Industries (LFS304-i) - AWS re:Invent 2018
Building IoT Devices for Regulated Industries (LFS304-i) - AWS re:Invent 2018
 
DevSecOps: Instituting Cultural Transformation for Public Sector Organization...
DevSecOps: Instituting Cultural Transformation for Public Sector Organization...DevSecOps: Instituting Cultural Transformation for Public Sector Organization...
DevSecOps: Instituting Cultural Transformation for Public Sector Organization...
 
Architecting for GxP Compliance in Life Sciences (LFS316) - AWS re:Invent 2018
Architecting for GxP Compliance in Life Sciences (LFS316) - AWS re:Invent 2018Architecting for GxP Compliance in Life Sciences (LFS316) - AWS re:Invent 2018
Architecting for GxP Compliance in Life Sciences (LFS316) - AWS re:Invent 2018
 
Detecting Financial Market Manipulation Using Machine Learning (AIM347) - AWS...
Detecting Financial Market Manipulation Using Machine Learning (AIM347) - AWS...Detecting Financial Market Manipulation Using Machine Learning (AIM347) - AWS...
Detecting Financial Market Manipulation Using Machine Learning (AIM347) - AWS...
 
Can Deep Learning Be Used to Treat Disease and Measure Product Value? (LFS310...
Can Deep Learning Be Used to Treat Disease and Measure Product Value? (LFS310...Can Deep Learning Be Used to Treat Disease and Measure Product Value? (LFS310...
Can Deep Learning Be Used to Treat Disease and Measure Product Value? (LFS310...
 
Build Deep Learning Applications Using PyTorch and Amazon SageMaker (AIM432-R...
Build Deep Learning Applications Using PyTorch and Amazon SageMaker (AIM432-R...Build Deep Learning Applications Using PyTorch and Amazon SageMaker (AIM432-R...
Build Deep Learning Applications Using PyTorch and Amazon SageMaker (AIM432-R...
 
Hollywood's Cloud-Based Content Lakes: Modernized Media Archives (MAE203) - A...
Hollywood's Cloud-Based Content Lakes: Modernized Media Archives (MAE203) - A...Hollywood's Cloud-Based Content Lakes: Modernized Media Archives (MAE203) - A...
Hollywood's Cloud-Based Content Lakes: Modernized Media Archives (MAE203) - A...
 
[NEW LAUNCH!] Introducing Amazon Forecast (AIM344) - AWS re:Invent 2018
[NEW LAUNCH!] Introducing Amazon Forecast  (AIM344) - AWS re:Invent 2018[NEW LAUNCH!] Introducing Amazon Forecast  (AIM344) - AWS re:Invent 2018
[NEW LAUNCH!] Introducing Amazon Forecast (AIM344) - AWS re:Invent 2018
 
Build a "Who's Who" App for Your Media Content (AIM409) - AWS re:Invent 2018
Build a "Who's Who" App for Your Media Content (AIM409) - AWS re:Invent 2018Build a "Who's Who" App for Your Media Content (AIM409) - AWS re:Invent 2018
Build a "Who's Who" App for Your Media Content (AIM409) - AWS re:Invent 2018
 
透過最新的 AWS 服務在 2019 年為您的業務轉型 (Level 200)
透過最新的 AWS 服務在 2019 年為您的業務轉型 (Level 200)透過最新的 AWS 服務在 2019 年為您的業務轉型 (Level 200)
透過最新的 AWS 服務在 2019 年為您的業務轉型 (Level 200)
 
Sequence-to-Sequence Modeling with Apache MXNet, Sockeye, and Amazon SageMake...
Sequence-to-Sequence Modeling with Apache MXNet, Sockeye, and Amazon SageMake...Sequence-to-Sequence Modeling with Apache MXNet, Sockeye, and Amazon SageMake...
Sequence-to-Sequence Modeling with Apache MXNet, Sockeye, and Amazon SageMake...
 
Build an Intelligent Multi-Modal User Agent with Voice and NLU (AIM340) - AWS...
Build an Intelligent Multi-Modal User Agent with Voice and NLU (AIM340) - AWS...Build an Intelligent Multi-Modal User Agent with Voice and NLU (AIM340) - AWS...
Build an Intelligent Multi-Modal User Agent with Voice and NLU (AIM340) - AWS...
 
Build an ETL Pipeline to Analyze Customer Data (AIM416) - AWS re:Invent 2018
Build an ETL Pipeline to Analyze Customer Data (AIM416) - AWS re:Invent 2018Build an ETL Pipeline to Analyze Customer Data (AIM416) - AWS re:Invent 2018
Build an ETL Pipeline to Analyze Customer Data (AIM416) - AWS re:Invent 2018
 
Deep Dive on Amazon S3 Storage Classes: Creating Cost Efficiencies across You...
Deep Dive on Amazon S3 Storage Classes: Creating Cost Efficiencies across You...Deep Dive on Amazon S3 Storage Classes: Creating Cost Efficiencies across You...
Deep Dive on Amazon S3 Storage Classes: Creating Cost Efficiencies across You...
 
AI/ML with Data Lakes: Counterintuitive Consumer Insights in Retail (RET206) ...
AI/ML with Data Lakes: Counterintuitive Consumer Insights in Retail (RET206) ...AI/ML with Data Lakes: Counterintuitive Consumer Insights in Retail (RET206) ...
AI/ML with Data Lakes: Counterintuitive Consumer Insights in Retail (RET206) ...
 
Go Fast and Remain Secure: How Millennium Enables Developers and Upholds Secu...
Go Fast and Remain Secure: How Millennium Enables Developers and Upholds Secu...Go Fast and Remain Secure: How Millennium Enables Developers and Upholds Secu...
Go Fast and Remain Secure: How Millennium Enables Developers and Upholds Secu...
 
Leadership Session: Innovation-Driven Healthcare (HLC201-iL) - AWS re:Invent ...
Leadership Session: Innovation-Driven Healthcare (HLC201-iL) - AWS re:Invent ...Leadership Session: Innovation-Driven Healthcare (HLC201-iL) - AWS re:Invent ...
Leadership Session: Innovation-Driven Healthcare (HLC201-iL) - AWS re:Invent ...
 

Similar to Natural Language Processing and Mortality Risk Prediction

Building and deploying AI/ML models on AWS for Biosciences professionals
Building and deploying AI/ML models on AWS for Biosciences professionalsBuilding and deploying AI/ML models on AWS for Biosciences professionals
Building and deploying AI/ML models on AWS for Biosciences professionalsjavier ramirez
 
NLP in Healthcare to Predict Adverse Events with Amazon SageMaker (AIM346) - ...
NLP in Healthcare to Predict Adverse Events with Amazon SageMaker (AIM346) - ...NLP in Healthcare to Predict Adverse Events with Amazon SageMaker (AIM346) - ...
NLP in Healthcare to Predict Adverse Events with Amazon SageMaker (AIM346) - ...Amazon Web Services
 
Enabling Sustainable Research Platforms in the Cloud
Enabling Sustainable Research Platforms in the CloudEnabling Sustainable Research Platforms in the Cloud
Enabling Sustainable Research Platforms in the CloudAmazon Web Services
 
Leadership Session: Accelerating Transformation in the Life Sciences (LFS201-...
Leadership Session: Accelerating Transformation in the Life Sciences (LFS201-...Leadership Session: Accelerating Transformation in the Life Sciences (LFS201-...
Leadership Session: Accelerating Transformation in the Life Sciences (LFS201-...Amazon Web Services
 
Drug Discovery Innovation in a Precompetitive Cloud Platform (LFS302-S) - AWS...
Drug Discovery Innovation in a Precompetitive Cloud Platform (LFS302-S) - AWS...Drug Discovery Innovation in a Precompetitive Cloud Platform (LFS302-S) - AWS...
Drug Discovery Innovation in a Precompetitive Cloud Platform (LFS302-S) - AWS...Amazon Web Services
 
Amazon SageMaker Ground Truth: Build High-Quality and Accurate ML Training Da...
Amazon SageMaker Ground Truth: Build High-Quality and Accurate ML Training Da...Amazon SageMaker Ground Truth: Build High-Quality and Accurate ML Training Da...
Amazon SageMaker Ground Truth: Build High-Quality and Accurate ML Training Da...Amazon Web Services
 
Empowering Every Brain! How Brain Power is using AWS-Powered AI in their Miss...
Empowering Every Brain! How Brain Power is using AWS-Powered AI in their Miss...Empowering Every Brain! How Brain Power is using AWS-Powered AI in their Miss...
Empowering Every Brain! How Brain Power is using AWS-Powered AI in their Miss...Amazon Web Services
 
Introducing Amazon SageMaker - AWS Online Tech Talks
Introducing Amazon SageMaker - AWS Online Tech TalksIntroducing Amazon SageMaker - AWS Online Tech Talks
Introducing Amazon SageMaker - AWS Online Tech TalksAmazon Web Services
 
Fraud Prevention and Detection on AWS
Fraud Prevention and Detection on AWSFraud Prevention and Detection on AWS
Fraud Prevention and Detection on AWSAmazon Web Services
 
Building a Recommender System on AWS
Building a Recommender System on AWSBuilding a Recommender System on AWS
Building a Recommender System on AWSAmazon Web Services
 
Datarobot, 자동화된 분석 적용 시 분석 절차의 변화 및 효용 - 홍운표 데이터 사이언티스트, DataRobot :: AWS Sum...
Datarobot, 자동화된 분석 적용 시 분석 절차의 변화 및 효용 - 홍운표 데이터 사이언티스트, DataRobot :: AWS Sum...Datarobot, 자동화된 분석 적용 시 분석 절차의 변화 및 효용 - 홍운표 데이터 사이언티스트, DataRobot :: AWS Sum...
Datarobot, 자동화된 분석 적용 시 분석 절차의 변화 및 효용 - 홍운표 데이터 사이언티스트, DataRobot :: AWS Sum...Amazon Web Services Korea
 
Harness the Power of Crowdsourcing with Amazon Mechanical Turk (AIM351) - AWS...
Harness the Power of Crowdsourcing with Amazon Mechanical Turk (AIM351) - AWS...Harness the Power of Crowdsourcing with Amazon Mechanical Turk (AIM351) - AWS...
Harness the Power of Crowdsourcing with Amazon Mechanical Turk (AIM351) - AWS...Amazon Web Services
 
AWS Public Sector Summit 2018, Data Supply Chain Pipeline
AWS Public Sector Summit 2018, Data Supply Chain PipelineAWS Public Sector Summit 2018, Data Supply Chain Pipeline
AWS Public Sector Summit 2018, Data Supply Chain PipelineStephen Moon
 
AWS Floor 28 - Migrating to AWS
AWS Floor 28 - Migrating to AWSAWS Floor 28 - Migrating to AWS
AWS Floor 28 - Migrating to AWSAdir Sharabi
 
AWS Community Day Nordics 2018 - Aino Health: Transition to serverless and le...
AWS Community Day Nordics 2018 - Aino Health: Transition to serverless and le...AWS Community Day Nordics 2018 - Aino Health: Transition to serverless and le...
AWS Community Day Nordics 2018 - Aino Health: Transition to serverless and le...Rolf Koski
 
Chaos Engineering and Scalability at Audible.com (ARC308) - AWS re:Invent 2018
Chaos Engineering and Scalability at Audible.com (ARC308) - AWS re:Invent 2018Chaos Engineering and Scalability at Audible.com (ARC308) - AWS re:Invent 2018
Chaos Engineering and Scalability at Audible.com (ARC308) - AWS re:Invent 2018Amazon Web Services
 
AI/ML Week: Improve Public Health
AI/ML Week: Improve Public HealthAI/ML Week: Improve Public Health
AI/ML Week: Improve Public HealthAmazon Web Services
 
Automated Retraining in SageMaker- AWS Startup Day Chicago 2018
Automated Retraining in SageMaker- AWS Startup Day Chicago 2018Automated Retraining in SageMaker- AWS Startup Day Chicago 2018
Automated Retraining in SageMaker- AWS Startup Day Chicago 2018Amazon Web Services
 
How Trupanion Became an AI-driven Company for Pets
How Trupanion Became an AI-driven Company for PetsHow Trupanion Became an AI-driven Company for Pets
How Trupanion Became an AI-driven Company for PetsAmazon Web Services
 
Chaos Engineering: Why Breaking Things Should Be Practiced - AWS Developer Wo...
Chaos Engineering: Why Breaking Things Should Be Practiced - AWS Developer Wo...Chaos Engineering: Why Breaking Things Should Be Practiced - AWS Developer Wo...
Chaos Engineering: Why Breaking Things Should Be Practiced - AWS Developer Wo...Amazon Web Services
 

Similar to Natural Language Processing and Mortality Risk Prediction (20)

Building and deploying AI/ML models on AWS for Biosciences professionals
Building and deploying AI/ML models on AWS for Biosciences professionalsBuilding and deploying AI/ML models on AWS for Biosciences professionals
Building and deploying AI/ML models on AWS for Biosciences professionals
 
NLP in Healthcare to Predict Adverse Events with Amazon SageMaker (AIM346) - ...
NLP in Healthcare to Predict Adverse Events with Amazon SageMaker (AIM346) - ...NLP in Healthcare to Predict Adverse Events with Amazon SageMaker (AIM346) - ...
NLP in Healthcare to Predict Adverse Events with Amazon SageMaker (AIM346) - ...
 
Enabling Sustainable Research Platforms in the Cloud
Enabling Sustainable Research Platforms in the CloudEnabling Sustainable Research Platforms in the Cloud
Enabling Sustainable Research Platforms in the Cloud
 
Leadership Session: Accelerating Transformation in the Life Sciences (LFS201-...
Leadership Session: Accelerating Transformation in the Life Sciences (LFS201-...Leadership Session: Accelerating Transformation in the Life Sciences (LFS201-...
Leadership Session: Accelerating Transformation in the Life Sciences (LFS201-...
 
Drug Discovery Innovation in a Precompetitive Cloud Platform (LFS302-S) - AWS...
Drug Discovery Innovation in a Precompetitive Cloud Platform (LFS302-S) - AWS...Drug Discovery Innovation in a Precompetitive Cloud Platform (LFS302-S) - AWS...
Drug Discovery Innovation in a Precompetitive Cloud Platform (LFS302-S) - AWS...
 
Amazon SageMaker Ground Truth: Build High-Quality and Accurate ML Training Da...
Amazon SageMaker Ground Truth: Build High-Quality and Accurate ML Training Da...Amazon SageMaker Ground Truth: Build High-Quality and Accurate ML Training Da...
Amazon SageMaker Ground Truth: Build High-Quality and Accurate ML Training Da...
 
Empowering Every Brain! How Brain Power is using AWS-Powered AI in their Miss...
Empowering Every Brain! How Brain Power is using AWS-Powered AI in their Miss...Empowering Every Brain! How Brain Power is using AWS-Powered AI in their Miss...
Empowering Every Brain! How Brain Power is using AWS-Powered AI in their Miss...
 
Introducing Amazon SageMaker - AWS Online Tech Talks
Introducing Amazon SageMaker - AWS Online Tech TalksIntroducing Amazon SageMaker - AWS Online Tech Talks
Introducing Amazon SageMaker - AWS Online Tech Talks
 
Fraud Prevention and Detection on AWS
Fraud Prevention and Detection on AWSFraud Prevention and Detection on AWS
Fraud Prevention and Detection on AWS
 
Building a Recommender System on AWS
Building a Recommender System on AWSBuilding a Recommender System on AWS
Building a Recommender System on AWS
 
Datarobot, 자동화된 분석 적용 시 분석 절차의 변화 및 효용 - 홍운표 데이터 사이언티스트, DataRobot :: AWS Sum...
Datarobot, 자동화된 분석 적용 시 분석 절차의 변화 및 효용 - 홍운표 데이터 사이언티스트, DataRobot :: AWS Sum...Datarobot, 자동화된 분석 적용 시 분석 절차의 변화 및 효용 - 홍운표 데이터 사이언티스트, DataRobot :: AWS Sum...
Datarobot, 자동화된 분석 적용 시 분석 절차의 변화 및 효용 - 홍운표 데이터 사이언티스트, DataRobot :: AWS Sum...
 
Harness the Power of Crowdsourcing with Amazon Mechanical Turk (AIM351) - AWS...
Harness the Power of Crowdsourcing with Amazon Mechanical Turk (AIM351) - AWS...Harness the Power of Crowdsourcing with Amazon Mechanical Turk (AIM351) - AWS...
Harness the Power of Crowdsourcing with Amazon Mechanical Turk (AIM351) - AWS...
 
AWS Public Sector Summit 2018, Data Supply Chain Pipeline
AWS Public Sector Summit 2018, Data Supply Chain PipelineAWS Public Sector Summit 2018, Data Supply Chain Pipeline
AWS Public Sector Summit 2018, Data Supply Chain Pipeline
 
AWS Floor 28 - Migrating to AWS
AWS Floor 28 - Migrating to AWSAWS Floor 28 - Migrating to AWS
AWS Floor 28 - Migrating to AWS
 
AWS Community Day Nordics 2018 - Aino Health: Transition to serverless and le...
AWS Community Day Nordics 2018 - Aino Health: Transition to serverless and le...AWS Community Day Nordics 2018 - Aino Health: Transition to serverless and le...
AWS Community Day Nordics 2018 - Aino Health: Transition to serverless and le...
 
Chaos Engineering and Scalability at Audible.com (ARC308) - AWS re:Invent 2018
Chaos Engineering and Scalability at Audible.com (ARC308) - AWS re:Invent 2018Chaos Engineering and Scalability at Audible.com (ARC308) - AWS re:Invent 2018
Chaos Engineering and Scalability at Audible.com (ARC308) - AWS re:Invent 2018
 
AI/ML Week: Improve Public Health
AI/ML Week: Improve Public HealthAI/ML Week: Improve Public Health
AI/ML Week: Improve Public Health
 
Automated Retraining in SageMaker- AWS Startup Day Chicago 2018
Automated Retraining in SageMaker- AWS Startup Day Chicago 2018Automated Retraining in SageMaker- AWS Startup Day Chicago 2018
Automated Retraining in SageMaker- AWS Startup Day Chicago 2018
 
How Trupanion Became an AI-driven Company for Pets
How Trupanion Became an AI-driven Company for PetsHow Trupanion Became an AI-driven Company for Pets
How Trupanion Became an AI-driven Company for Pets
 
Chaos Engineering: Why Breaking Things Should Be Practiced - AWS Developer Wo...
Chaos Engineering: Why Breaking Things Should Be Practiced - AWS Developer Wo...Chaos Engineering: Why Breaking Things Should Be Practiced - AWS Developer Wo...
Chaos Engineering: Why Breaking Things Should Be Practiced - AWS Developer Wo...
 

More from Amazon Web Services

Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...
Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...
Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...Amazon Web Services
 
Big Data per le Startup: come creare applicazioni Big Data in modalità Server...
Big Data per le Startup: come creare applicazioni Big Data in modalità Server...Big Data per le Startup: come creare applicazioni Big Data in modalità Server...
Big Data per le Startup: come creare applicazioni Big Data in modalità Server...Amazon Web Services
 
Esegui pod serverless con Amazon EKS e AWS Fargate
Esegui pod serverless con Amazon EKS e AWS FargateEsegui pod serverless con Amazon EKS e AWS Fargate
Esegui pod serverless con Amazon EKS e AWS FargateAmazon Web Services
 
Costruire Applicazioni Moderne con AWS
Costruire Applicazioni Moderne con AWSCostruire Applicazioni Moderne con AWS
Costruire Applicazioni Moderne con AWSAmazon Web Services
 
Come spendere fino al 90% in meno con i container e le istanze spot
Come spendere fino al 90% in meno con i container e le istanze spot Come spendere fino al 90% in meno con i container e le istanze spot
Come spendere fino al 90% in meno con i container e le istanze spot Amazon Web Services
 
Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...
Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...
Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...Amazon Web Services
 
OpsWorks Configuration Management: automatizza la gestione e i deployment del...
OpsWorks Configuration Management: automatizza la gestione e i deployment del...OpsWorks Configuration Management: automatizza la gestione e i deployment del...
OpsWorks Configuration Management: automatizza la gestione e i deployment del...Amazon Web Services
 
Microsoft Active Directory su AWS per supportare i tuoi Windows Workloads
Microsoft Active Directory su AWS per supportare i tuoi Windows WorkloadsMicrosoft Active Directory su AWS per supportare i tuoi Windows Workloads
Microsoft Active Directory su AWS per supportare i tuoi Windows WorkloadsAmazon Web Services
 
Database Oracle e VMware Cloud on AWS i miti da sfatare
Database Oracle e VMware Cloud on AWS i miti da sfatareDatabase Oracle e VMware Cloud on AWS i miti da sfatare
Database Oracle e VMware Cloud on AWS i miti da sfatareAmazon Web Services
 
Crea la tua prima serverless ledger-based app con QLDB e NodeJS
Crea la tua prima serverless ledger-based app con QLDB e NodeJSCrea la tua prima serverless ledger-based app con QLDB e NodeJS
Crea la tua prima serverless ledger-based app con QLDB e NodeJSAmazon Web Services
 
API moderne real-time per applicazioni mobili e web
API moderne real-time per applicazioni mobili e webAPI moderne real-time per applicazioni mobili e web
API moderne real-time per applicazioni mobili e webAmazon Web Services
 
Database Oracle e VMware Cloud™ on AWS: i miti da sfatare
Database Oracle e VMware Cloud™ on AWS: i miti da sfatareDatabase Oracle e VMware Cloud™ on AWS: i miti da sfatare
Database Oracle e VMware Cloud™ on AWS: i miti da sfatareAmazon Web Services
 
Tools for building your MVP on AWS
Tools for building your MVP on AWSTools for building your MVP on AWS
Tools for building your MVP on AWSAmazon Web Services
 
How to Build a Winning Pitch Deck
How to Build a Winning Pitch DeckHow to Build a Winning Pitch Deck
How to Build a Winning Pitch DeckAmazon Web Services
 
Building a web application without servers
Building a web application without serversBuilding a web application without servers
Building a web application without serversAmazon Web Services
 
AWS_HK_StartupDay_Building Interactive websites while automating for efficien...
AWS_HK_StartupDay_Building Interactive websites while automating for efficien...AWS_HK_StartupDay_Building Interactive websites while automating for efficien...
AWS_HK_StartupDay_Building Interactive websites while automating for efficien...Amazon Web Services
 
Introduzione a Amazon Elastic Container Service
Introduzione a Amazon Elastic Container ServiceIntroduzione a Amazon Elastic Container Service
Introduzione a Amazon Elastic Container ServiceAmazon Web Services
 

More from Amazon Web Services (20)

Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...
Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...
Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...
 
Big Data per le Startup: come creare applicazioni Big Data in modalità Server...
Big Data per le Startup: come creare applicazioni Big Data in modalità Server...Big Data per le Startup: come creare applicazioni Big Data in modalità Server...
Big Data per le Startup: come creare applicazioni Big Data in modalità Server...
 
Esegui pod serverless con Amazon EKS e AWS Fargate
Esegui pod serverless con Amazon EKS e AWS FargateEsegui pod serverless con Amazon EKS e AWS Fargate
Esegui pod serverless con Amazon EKS e AWS Fargate
 
Costruire Applicazioni Moderne con AWS
Costruire Applicazioni Moderne con AWSCostruire Applicazioni Moderne con AWS
Costruire Applicazioni Moderne con AWS
 
Come spendere fino al 90% in meno con i container e le istanze spot
Come spendere fino al 90% in meno con i container e le istanze spot Come spendere fino al 90% in meno con i container e le istanze spot
Come spendere fino al 90% in meno con i container e le istanze spot
 
Open banking as a service
Open banking as a serviceOpen banking as a service
Open banking as a service
 
Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...
Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...
Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...
 
OpsWorks Configuration Management: automatizza la gestione e i deployment del...
OpsWorks Configuration Management: automatizza la gestione e i deployment del...OpsWorks Configuration Management: automatizza la gestione e i deployment del...
OpsWorks Configuration Management: automatizza la gestione e i deployment del...
 
Microsoft Active Directory su AWS per supportare i tuoi Windows Workloads
Microsoft Active Directory su AWS per supportare i tuoi Windows WorkloadsMicrosoft Active Directory su AWS per supportare i tuoi Windows Workloads
Microsoft Active Directory su AWS per supportare i tuoi Windows Workloads
 
Computer Vision con AWS
Computer Vision con AWSComputer Vision con AWS
Computer Vision con AWS
 
Database Oracle e VMware Cloud on AWS i miti da sfatare
Database Oracle e VMware Cloud on AWS i miti da sfatareDatabase Oracle e VMware Cloud on AWS i miti da sfatare
Database Oracle e VMware Cloud on AWS i miti da sfatare
 
Crea la tua prima serverless ledger-based app con QLDB e NodeJS
Crea la tua prima serverless ledger-based app con QLDB e NodeJSCrea la tua prima serverless ledger-based app con QLDB e NodeJS
Crea la tua prima serverless ledger-based app con QLDB e NodeJS
 
API moderne real-time per applicazioni mobili e web
API moderne real-time per applicazioni mobili e webAPI moderne real-time per applicazioni mobili e web
API moderne real-time per applicazioni mobili e web
 
Database Oracle e VMware Cloud™ on AWS: i miti da sfatare
Database Oracle e VMware Cloud™ on AWS: i miti da sfatareDatabase Oracle e VMware Cloud™ on AWS: i miti da sfatare
Database Oracle e VMware Cloud™ on AWS: i miti da sfatare
 
Tools for building your MVP on AWS
Tools for building your MVP on AWSTools for building your MVP on AWS
Tools for building your MVP on AWS
 
How to Build a Winning Pitch Deck
How to Build a Winning Pitch DeckHow to Build a Winning Pitch Deck
How to Build a Winning Pitch Deck
 
Building a web application without servers
Building a web application without serversBuilding a web application without servers
Building a web application without servers
 
Fundraising Essentials
Fundraising EssentialsFundraising Essentials
Fundraising Essentials
 
AWS_HK_StartupDay_Building Interactive websites while automating for efficien...
AWS_HK_StartupDay_Building Interactive websites while automating for efficien...AWS_HK_StartupDay_Building Interactive websites while automating for efficien...
AWS_HK_StartupDay_Building Interactive websites while automating for efficien...
 
Introduzione a Amazon Elastic Container Service
Introduzione a Amazon Elastic Container ServiceIntroduzione a Amazon Elastic Container Service
Introduzione a Amazon Elastic Container Service
 

Natural Language Processing and Mortality Risk Prediction

  • 1.
  • 2. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. Ujjwal Ratan Principal AI/ML SA, GLS AWS L F S 3 0 3 Mengqi Jin Machine Learning Scientist AWS Bryan Coon Sr. Manager, Solutions Architecture, GLS AWS Natural Language Processing and Mortality Risk Prediction Using Amazon Comprehend Medical & Amazon SageMaker
  • 3. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. Agenda What is Amazon Comprehend Medical? What is Amazon SageMaker? Workshop use case Workshop architecture Hands on
  • 4. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
  • 5. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. Amazon Comprehend Medical A new HIPAA-eligible service that uses machine learning to extract medical information with high accuracy, reducing the cost, time and effort of processing large amounts of unstructured medical text
  • 6. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. Amazon Comprehend Medical Entities • Medication • Medical condition • Test, Treatments and Procedures • Anatomy • Protected Health Information (PHI) Relationship Extraction • Medication and dosage • Test and result • Many more Entity Traits • Negation • Diagnosis, Sign or Symptom Protected Health Information Identification (PHId API) Distill a complex process into a simple API call Medical Named Entity and Relationship Extraction (NERe API)
  • 7. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. Protected Health Information (PHI) Mr. Doe : Name 63 : Age Comprehend Medical - NERe API Mr. Doe is a 63-year-old gentleman with coronary artery disease and hypertension. CURRENT MEDICATIONS: taking a dose of LIPITOR 20 mg once daily. aws comprehend-medical detect-entities --region us-east-1 --text “<Insert Text Here>” Named Entities
  • 8. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. Comprehend Medical – NERe API CURRENT MEDICATIONS: taking a dose of LIPITOR 20 mg once daily aws comprehend-medical detect-entities --region us-east-1 --text “<Insert Text Here>” Relationship Extraction
  • 9. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. Amazon Comprehend – NERe API What are Entity Traits? (1) Negation (2) Diagnosis, sign or symptom Rash Discontinue Abraxane, patient denies taking Tylenol 325 mg and is not taking calcium carbonate, patient also stopped taking cholecalciferol 1,000 units PO Entity Traits
  • 10. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. Amazon Comprehend – PHId API Mr. Doe is a 63-year-old gentleman with coronary artery disease and hypertension. He currently lives in Anytown and works as a teacher. He currently lives in Anytown and works as a teacher. His PCP, Dr. Stiles, works at the University of Anystate. aws comprehend-medical detect-phi --region us- east-1 --text “<Insert Text Here>” In addition to extracting PHI, the PHId API identifies relevant patient identifiers described in HIPAA Safe Harbor method of de-identification Personal Health Information
  • 11. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
  • 12. “For cancer patients and the researchers dedicated to curing them, time is the limiting resource. The process of developing clinical trials and connecting them with the right patients requires research teams to sift through and label mountains of clinical record data. The Amazon Comprehend Medical service reduces this time burden from hours to seconds. This is a vital step toward getting researchers rapid access to the information they need when they need it so they can advance lifesaving therapies for patients.” Matthew Trunnell, Chief Information Officer, Fred Hutchinson Cancer Research Center
  • 13. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
  • 14. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. A fully managed service that enables data scientists and developers to quickly and easily build machine-learning based models into production smart applications. Amazon SageMaker
  • 15. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. End-to-End Machine Learning Platform Zero setup Flexible Model Training Pay by the second Amazon SageMaker Build, train, and deploy machine learning models at scale
  • 16. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. One-click training for ML, DL, and custom algorithms Easier training with hyperparameter optimization Highly optimized machine learning algorithms Deployment without engineering effort Fully managed hosting at scale BuildPre-built notebook instances Deploy Train Amazon SageMaker
  • 17. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
  • 18. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. Business problem Predict an ICU patient’s risk of mortality Why is this important?
  • 19. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
  • 20. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. This workshop has four sections
  • 21. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
  • 22. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. Go to the following link to begin setup https://bit.ly/2KcWj6R
  • 23. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. Architecture diagram
  • 24. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. Mortality risk prediction Amazon ECR Amazon SageMaker Amazon S3 Data Algorithm Training Inference Model Artifacts Prediction Outputs Amazon S3 Doc2Vec Note Embeddings 1 32 Notebooks
  • 25. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
  • 26. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. Mortality risk prediction 0 hr 12 hr 24 hr 36 hr 48 hr … … … … Discharge Observation Window Prediction Window Death? DischargeICU Admission Prediction • Goal: Early detection of at-risk patients in ICU • Using first 48 hrs data of admission to predict in-hospital mortality • Numerical Variables: • Vital Signs • Labs • Bedside Measurements • Unstructured Notes: • Nursing Notes • Physician Notes • ECG Notes • Other 12 types of notes Mortality is one of the metrics used by CMS to evaluate quality of care and hospital ranking
  • 27. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. Risk modeling with text information Multimodal Embedding Raw Notes Model Artifacts Doc2VecC Medication Diagnosis Treatment/ Test/ Procedure Recurrent Neural Network Amazon Comprehend Medical Extracted Medical Entities Concatenated entities Step 1: Text Processing Step 2: Training Step 3: Offline Testing Evaluation Metrics
  • 28. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. Risk modeling with text information Step 1: Generate Document Embeddings • Use Doc2VecC to generate embeddings for 4000 medical notes • Save trained embedding dictionary on Amazon Simple Storage Service (Amazon S3) Input Raw notes saved in csv files Output NumPy dictionary of document embeddings keyed by Note ID
  • 29. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. Risk modeling with text information Step 2: Amazon SageMaker Training Job • Run the algorithm over the training data • Save the output model and select the best one to save it on Amazon S3 Input Time series data Pre-trained document embeddings Output Model artifacts saved in model.tar.gz file
  • 30. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. Risk modeling with text information Step 3: Amazon SageMaker BatchTransform Job • Making inference on test data set • Evaluate model performance Input Selected best model saved in model.tar.gz file Time series data Pre-trained document embeddings Output Prediction for each time series record Evaluation metrics
  • 31. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
  • 32. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
  • 33. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. Related breakouts Thursday November 29th Introducing Amazon Comprehend Medical 12:15 – 1:15 PM | Aria West Level 3 Juniper 4
  • 34. Thank you! © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved. Ujjwal Ratan ujjwalr@amazon.com Bryan Coon bccn@amazon.com Mengqi Jin maggjin@amazon.com
  • 35. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.