SlideShare a Scribd company logo
Computational Classification
Techniques for Neuroimaging
A Machine Learning Based Approach
Adrian Smith – Undergraduate
Computer Science Department
Sonoma State University
Fundamentals
• Understanding the human
brain has been a central
theme of human history
• By growing our
understanding of the brain,
we improve our ability to
treat diseases (Gur2002)
• Understanding the brain
helps us be aware of it’s
limitations
Artist’s Depiction of Neurons
UCI Research
Courtesy of OSA Student Chapter at UCI Art in Science Contest.
Photo by: Ardy Rahman
fMRI Scanning
• Functional Magnetic
Resonance Imaging (fMRI)
allows us to measure localized
brain activity
• This allows one to find
relationships between cognition
and brain activity
• Blood oxygen is used as a
measure of activity (BOLD
imaging)
• This technique produces rich
data, but contains high levels of
noise
CSRB (Keck MRI Center)
Data Collection
• One major advantage of
researching fMRI data is it’s
availability on a variety of
online locations
• We worked with 1452 total
brain scans each
corresponding to one of 9
categories
• The categories refer to the
image a subject was
observing
Analysis Goals
• Our goal was to be able to, given the
fMRI scan of a subject, predict what
image they were observing
• This means differentiating scans
based on the image the subject is
observing
• What is the relationship?
Haxby2001 Stimulus Images
Machine Learning Techniques
• Machine learning is an
information processing
technique
• The field of machine learning is
at the heart of understanding
“Big Data”
• We aimed to use modern
machine learning techniques to
help classify fMRI data.
How does Machine Learning Work?
• Machine Learning classification
focuses on designing
algorithms which are trained to
categorize objects
• This is done by combining
some defining characteristics
and a label
• The algorithm trains on one set
of data, and then is tested to
see how accurately it can
predict the label of some piece
of data.
• What is the data?
By Antti Ajanki AnAj (Own work) [GFDL
(http://www.gnu.org/copyleft/fdl.html), CC-BY-SA-
3.0 (http://creativecommons.org/licenses/by-sa/3.0/)
Which is active before processing?
Unprocessed Active Unprocessed Rest
Which is active after processing?
Processed Active Processed Rest
Preprocessing
• We applied masks that came with
the dataset in order to focus on the
Ventral Temporal cortex, our region
of interest
• We then applied a polynomial
detrender, which eliminates
systematic trends, such as signal
increase as the machine warms up
• This was followed by a key step, z-
scoring against the rest position
Graph of Normal Distribution
Public Domain
Classification
• We now had to decide how to
process the image data
• This meant choosing features
that best represented the data
we sought
• We also tested a variety of
classification algorithms which
would label images based on the
chosen feature
Features
• We started with the our preprocessed values, and then looked at a
variety of transforms
• We chose the full vector and the PCA reduced version as our main
features of interest
• Principle Component Analysis (PCA) is a tool to reduce the dimensionality of a
dataset
PCA
Full Vector (Samples)
50 Highest Values
Histogram
[0.5, .01, -.02, 1.5, 2.0, … -3.0]576
One Volume
Experimental Design
• Data was split evenly and randomly into training and test
• We used several feature vectors to test each classifier
• We primarily focused on k Nearest Neighbor (kNN) and Support
Vector Machine (SVM) classifiers
• Tests were repeated 15 times and scores averaged
Train
Feature
Training
Label
Trained
Classifier
Testing
Feature
Predicted
Label
Testing
Label
Comparison
Accuracy and
Confusion
Matrix
Classifier
kNN vs. SVM
• SVM preforms better than kNN
• Increase in accuracy is likely due to the weakness of kNN when
dealing with high dimensionality
SVM on samples, 90.9%
accuracy
kNN on samples, 75.6%
accuracy
• We applied PCA to the processed data
• This produced a vector over half the size of our original
• This smaller vector produces more accurate results
Samples vs. PCA
PCA (SVM), 92.1%
accuracy
SVM on samples, 90.9%
accuracy
• PCA and SVM in combination gave the best results
after repeated testing
• We achieved on average 92.1% accuracy among 9
labels, with a 2.0% standard deviation.
• Our classification methods are effective and
repeatable
• We also gained a variety of insights about the
nature of the data
Classification Results: Accuracy
• We saw several labels which
repeatedly misclassified, and
saw accuracy improve as they
were removed
• One area of further study is
investigating whether these
patterns exist between multiple
subjects, and why
PCA (SVM), 92.1%
accuracy
Classification Results: Insights
Future Exploration
• We intend to move towards classifying
across multiple subjects
• This is of utmost importance to clinical
applications of fMRI data
• Multisubject comparison presents
challenges due to the variation in brain
structure
• We intend to build upon previous work on
feature detection and scaling maps
(Gill2014)
Sources
• Gur, R. E., McGrath, C., Chan, R. M., Schroeder, L., Turner, T., Turetsky, B. I., ...
& Gur, R. C. (2002). An fMRI study of facial emotion processing in patients with
schizophrenia. American Journal of Psychiatry.
• Haxby, J. V., Gobbini, M. I., Furey, M. L., Ishai, A., Schouten, J. L., & Pietrini, P.
(2001). Distributed and overlapping representations of faces and objects in
ventral temporal cortex. Science, 293(5539), 2425-2430.
• Gill, G., Bauer, C., & Beichel, R. R. (2014). A method for avoiding overlap of left
and right lungs in shape model guided segmentation of lungs in CT volumes.
Medical physics, 41(10), 101908.
• Dataset: This data was obtained from the OpenfMRI database. Its accession
number is ds000105. The original authors of :ref:`Haxby et al. (2001) <HGF+01>`
hold the copyright of this dataset and made it available under the terms of the
`Creative Commons Attribution-Share Alike 3.0`_ license.
Acknowledgments
• Dr. Gurman Gill – Mentor
• OpenfMRI – Source of all data, and amazing example of open
data in science
• pyMVPA – Python toolkit used in preprocessing
• Scikit-learn – Python toolkit used in classification
• Dr. Yaroslav Halchenko – Researcher who provided extensive
aid in understanding and dealing with fMRI data
Questions?
Extra Graphics
SVM of top 400 values.
30.9% accuracy SVM on 90% PCA. 92.2%
accuracy

More Related Content

What's hot

IRJET - Deep Learning Approach to Inpainting and Outpainting System
IRJET -  	  Deep Learning Approach to Inpainting and Outpainting SystemIRJET -  	  Deep Learning Approach to Inpainting and Outpainting System
IRJET - Deep Learning Approach to Inpainting and Outpainting System
IRJET Journal
 
My experiment
My experimentMy experiment
My experiment
Boshra Albayaty
 
MIP AND UNSUPERVISED CLUSTERING FOR THE DETECTION OF BRAIN TUMOUR CELLS
MIP AND UNSUPERVISED CLUSTERING FOR THE DETECTION OF BRAIN TUMOUR CELLSMIP AND UNSUPERVISED CLUSTERING FOR THE DETECTION OF BRAIN TUMOUR CELLS
MIP AND UNSUPERVISED CLUSTERING FOR THE DETECTION OF BRAIN TUMOUR CELLS
AM Publications
 
Comparison between the genetic algorithms optimization and particle swarm opt...
Comparison between the genetic algorithms optimization and particle swarm opt...Comparison between the genetic algorithms optimization and particle swarm opt...
Comparison between the genetic algorithms optimization and particle swarm opt...
IAEME Publication
 
Clustering of medline documents using semi supervised spectral clustering
Clustering of medline documents using semi supervised spectral clusteringClustering of medline documents using semi supervised spectral clustering
Clustering of medline documents using semi supervised spectral clustering
eSAT Journals
 
Segmentation of unhealthy region of plant leaf using image processing techniques
Segmentation of unhealthy region of plant leaf using image processing techniquesSegmentation of unhealthy region of plant leaf using image processing techniques
Segmentation of unhealthy region of plant leaf using image processing techniques
eSAT Journals
 
A NOVEL APPROACH FOR FEATURE EXTRACTION AND SELECTION ON MRI IMAGES FOR BRAIN...
A NOVEL APPROACH FOR FEATURE EXTRACTION AND SELECTION ON MRI IMAGES FOR BRAIN...A NOVEL APPROACH FOR FEATURE EXTRACTION AND SELECTION ON MRI IMAGES FOR BRAIN...
A NOVEL APPROACH FOR FEATURE EXTRACTION AND SELECTION ON MRI IMAGES FOR BRAIN...
cscpconf
 
Two are better than one IEEE-SMC talk
Two are better than one IEEE-SMC talkTwo are better than one IEEE-SMC talk
Two are better than one IEEE-SMC talk
Kyongsik Yun
 
New Feature Selection Model Based Ensemble Rule Classifiers Method for Datase...
New Feature Selection Model Based Ensemble Rule Classifiers Method for Datase...New Feature Selection Model Based Ensemble Rule Classifiers Method for Datase...
New Feature Selection Model Based Ensemble Rule Classifiers Method for Datase...
ijaia
 
Are we really including all relevant evidence
Are we really including all relevant evidence Are we really including all relevant evidence
Are we really including all relevant evidence
cheweb1
 
Deep Attention Model for Triage of Emergency Department Patients - Djordje Gl...
Deep Attention Model for Triage of Emergency Department Patients - Djordje Gl...Deep Attention Model for Triage of Emergency Department Patients - Djordje Gl...
Deep Attention Model for Triage of Emergency Department Patients - Djordje Gl...
Institute of Contemporary Sciences
 
Mapping to the Metabolomic Manifold
Mapping to the Metabolomic ManifoldMapping to the Metabolomic Manifold
Mapping to the Metabolomic Manifold
Dmitry Grapov
 
Temporal based Recommendation System
Temporal based Recommendation SystemTemporal based Recommendation System
Temporal based Recommendation System
Nurfadhlina Mohd Sharef
 
Identification of Disease in Leaves using Genetic Algorithm
Identification of Disease in Leaves using Genetic AlgorithmIdentification of Disease in Leaves using Genetic Algorithm
Identification of Disease in Leaves using Genetic Algorithm
ijtsrd
 
Hanaa phd presentation 14-4-2017
Hanaa phd  presentation  14-4-2017Hanaa phd  presentation  14-4-2017
Hanaa phd presentation 14-4-2017
Aboul Ella Hassanien
 
Plant Leaf Disease Analysis using Image Processing Technique with Modified SV...
Plant Leaf Disease Analysis using Image Processing Technique with Modified SV...Plant Leaf Disease Analysis using Image Processing Technique with Modified SV...
Plant Leaf Disease Analysis using Image Processing Technique with Modified SV...
Tarun Kumar
 
Parkinson disease classification recorded v2.0
Parkinson disease classification recorded   v2.0Parkinson disease classification recorded   v2.0
Parkinson disease classification recorded v2.0
Nikhil Shrivastava, MS, SAFe PMPO
 
Connecting Metabolomic Data with Context
Connecting Metabolomic Data with ContextConnecting Metabolomic Data with Context
Connecting Metabolomic Data with Context
Dmitry Grapov
 

What's hot (20)

IRJET - Deep Learning Approach to Inpainting and Outpainting System
IRJET -  	  Deep Learning Approach to Inpainting and Outpainting SystemIRJET -  	  Deep Learning Approach to Inpainting and Outpainting System
IRJET - Deep Learning Approach to Inpainting and Outpainting System
 
My experiment
My experimentMy experiment
My experiment
 
MIP AND UNSUPERVISED CLUSTERING FOR THE DETECTION OF BRAIN TUMOUR CELLS
MIP AND UNSUPERVISED CLUSTERING FOR THE DETECTION OF BRAIN TUMOUR CELLSMIP AND UNSUPERVISED CLUSTERING FOR THE DETECTION OF BRAIN TUMOUR CELLS
MIP AND UNSUPERVISED CLUSTERING FOR THE DETECTION OF BRAIN TUMOUR CELLS
 
Comparison between the genetic algorithms optimization and particle swarm opt...
Comparison between the genetic algorithms optimization and particle swarm opt...Comparison between the genetic algorithms optimization and particle swarm opt...
Comparison between the genetic algorithms optimization and particle swarm opt...
 
Clustering of medline documents using semi supervised spectral clustering
Clustering of medline documents using semi supervised spectral clusteringClustering of medline documents using semi supervised spectral clustering
Clustering of medline documents using semi supervised spectral clustering
 
Segmentation of unhealthy region of plant leaf using image processing techniques
Segmentation of unhealthy region of plant leaf using image processing techniquesSegmentation of unhealthy region of plant leaf using image processing techniques
Segmentation of unhealthy region of plant leaf using image processing techniques
 
A NOVEL APPROACH FOR FEATURE EXTRACTION AND SELECTION ON MRI IMAGES FOR BRAIN...
A NOVEL APPROACH FOR FEATURE EXTRACTION AND SELECTION ON MRI IMAGES FOR BRAIN...A NOVEL APPROACH FOR FEATURE EXTRACTION AND SELECTION ON MRI IMAGES FOR BRAIN...
A NOVEL APPROACH FOR FEATURE EXTRACTION AND SELECTION ON MRI IMAGES FOR BRAIN...
 
Two are better than one IEEE-SMC talk
Two are better than one IEEE-SMC talkTwo are better than one IEEE-SMC talk
Two are better than one IEEE-SMC talk
 
New Feature Selection Model Based Ensemble Rule Classifiers Method for Datase...
New Feature Selection Model Based Ensemble Rule Classifiers Method for Datase...New Feature Selection Model Based Ensemble Rule Classifiers Method for Datase...
New Feature Selection Model Based Ensemble Rule Classifiers Method for Datase...
 
Are we really including all relevant evidence
Are we really including all relevant evidence Are we really including all relevant evidence
Are we really including all relevant evidence
 
Deep Attention Model for Triage of Emergency Department Patients - Djordje Gl...
Deep Attention Model for Triage of Emergency Department Patients - Djordje Gl...Deep Attention Model for Triage of Emergency Department Patients - Djordje Gl...
Deep Attention Model for Triage of Emergency Department Patients - Djordje Gl...
 
Mapping to the Metabolomic Manifold
Mapping to the Metabolomic ManifoldMapping to the Metabolomic Manifold
Mapping to the Metabolomic Manifold
 
Temporal based Recommendation System
Temporal based Recommendation SystemTemporal based Recommendation System
Temporal based Recommendation System
 
Identification of Disease in Leaves using Genetic Algorithm
Identification of Disease in Leaves using Genetic AlgorithmIdentification of Disease in Leaves using Genetic Algorithm
Identification of Disease in Leaves using Genetic Algorithm
 
FYP
FYPFYP
FYP
 
Hanaa phd presentation 14-4-2017
Hanaa phd  presentation  14-4-2017Hanaa phd  presentation  14-4-2017
Hanaa phd presentation 14-4-2017
 
Plant Leaf Disease Analysis using Image Processing Technique with Modified SV...
Plant Leaf Disease Analysis using Image Processing Technique with Modified SV...Plant Leaf Disease Analysis using Image Processing Technique with Modified SV...
Plant Leaf Disease Analysis using Image Processing Technique with Modified SV...
 
Parkinson disease classification recorded v2.0
Parkinson disease classification recorded   v2.0Parkinson disease classification recorded   v2.0
Parkinson disease classification recorded v2.0
 
Connecting Metabolomic Data with Context
Connecting Metabolomic Data with ContextConnecting Metabolomic Data with Context
Connecting Metabolomic Data with Context
 
Technical Portion of PhD Research
Technical Portion of PhD ResearchTechnical Portion of PhD Research
Technical Portion of PhD Research
 

Viewers also liked

සතියේ සුපුවත july 26 aug 01
සතියේ සුපුවත  july 26 aug 01සතියේ සුපුවත  july 26 aug 01
සතියේ සුපුවත july 26 aug 01
Adasupuwatha2
 
Certificate Saurabh Bhargava
Certificate Saurabh BhargavaCertificate Saurabh Bhargava
Certificate Saurabh BhargavaSAURABH BHARGAVA
 
Maestra virutdes. el mejor aestudiante
Maestra virutdes. el mejor aestudianteMaestra virutdes. el mejor aestudiante
Maestra virutdes. el mejor aestudianteMineducyt El Salvador
 
Jornadas ul
Jornadas ulJornadas ul
Jornadas ul
Maria Pereira
 
Yellow Shade Wrinkle-Resistant Background
Yellow Shade Wrinkle-Resistant BackgroundYellow Shade Wrinkle-Resistant Background
Yellow Shade Wrinkle-Resistant Background
neville ocassidy
 
Westlaw Insight (Deferred Prosecution Agreements)
Westlaw Insight (Deferred Prosecution Agreements)Westlaw Insight (Deferred Prosecution Agreements)
Westlaw Insight (Deferred Prosecution Agreements)Kathryn Hughes
 
1999CertifTitulo.Ing.Quimico
1999CertifTitulo.Ing.Quimico1999CertifTitulo.Ing.Quimico
1999CertifTitulo.Ing.QuimicoJorge Hugo Meza V
 
Aleix xisco informatica
Aleix xisco informaticaAleix xisco informatica
Aleix xisco informaticaaleix6
 
Meet anna rose pierre
Meet anna rose pierreMeet anna rose pierre
Meet anna rose pierre
pierreannarose
 

Viewers also liked (15)

Diagram-1
Diagram-1Diagram-1
Diagram-1
 
Storm Sewer Design Cert
Storm Sewer Design CertStorm Sewer Design Cert
Storm Sewer Design Cert
 
Http
HttpHttp
Http
 
සතියේ සුපුවත july 26 aug 01
සතියේ සුපුවත  july 26 aug 01සතියේ සුපුවත  july 26 aug 01
සතියේ සුපුවත july 26 aug 01
 
ประวัติส่วนตัว
ประวัติส่วนตัวประวัติส่วนตัว
ประวัติส่วนตัว
 
Certificate Saurabh Bhargava
Certificate Saurabh BhargavaCertificate Saurabh Bhargava
Certificate Saurabh Bhargava
 
Maestra virutdes. el mejor aestudiante
Maestra virutdes. el mejor aestudianteMaestra virutdes. el mejor aestudiante
Maestra virutdes. el mejor aestudiante
 
Pos-DOC
Pos-DOCPos-DOC
Pos-DOC
 
Jornadas ul
Jornadas ulJornadas ul
Jornadas ul
 
Yellow Shade Wrinkle-Resistant Background
Yellow Shade Wrinkle-Resistant BackgroundYellow Shade Wrinkle-Resistant Background
Yellow Shade Wrinkle-Resistant Background
 
Family Ski Photo
Family Ski PhotoFamily Ski Photo
Family Ski Photo
 
Westlaw Insight (Deferred Prosecution Agreements)
Westlaw Insight (Deferred Prosecution Agreements)Westlaw Insight (Deferred Prosecution Agreements)
Westlaw Insight (Deferred Prosecution Agreements)
 
1999CertifTitulo.Ing.Quimico
1999CertifTitulo.Ing.Quimico1999CertifTitulo.Ing.Quimico
1999CertifTitulo.Ing.Quimico
 
Aleix xisco informatica
Aleix xisco informaticaAleix xisco informatica
Aleix xisco informatica
 
Meet anna rose pierre
Meet anna rose pierreMeet anna rose pierre
Meet anna rose pierre
 

Similar to CSU_comp

Automatic System for Detection and Classification of Brain Tumors
Automatic System for Detection and Classification of Brain TumorsAutomatic System for Detection and Classification of Brain Tumors
Automatic System for Detection and Classification of Brain Tumors
Fatma Sayed Ibrahim
 
Diagnosis Support by Machine Learning Using Posturography Data
Diagnosis Support by Machine Learning Using Posturography DataDiagnosis Support by Machine Learning Using Posturography Data
Diagnosis Support by Machine Learning Using Posturography Data
TeruKamogashira
 
Machine Learning and Deep Contemplation of Data
Machine Learning and Deep Contemplation of DataMachine Learning and Deep Contemplation of Data
Machine Learning and Deep Contemplation of Data
Joel Saltz
 
Artificial Intelligence in pathology
Artificial Intelligence in pathologyArtificial Intelligence in pathology
Artificial Intelligence in pathology
nehaSingh1543
 
MRI Brain Tumour Classification Using SURF and SIFT Features
MRI Brain Tumour Classification Using SURF and SIFT FeaturesMRI Brain Tumour Classification Using SURF and SIFT Features
MRI Brain Tumour Classification Using SURF and SIFT Features
IJMTST Journal
 
A Review on the Brain Tumor Detection and Segmentation Techniques
A Review on the Brain Tumor Detection and Segmentation TechniquesA Review on the Brain Tumor Detection and Segmentation Techniques
A Review on the Brain Tumor Detection and Segmentation Techniques
IRJET Journal
 
braintumordetectionusingimagesegmentationppt-210830184640 (1).pdf
braintumordetectionusingimagesegmentationppt-210830184640 (1).pdfbraintumordetectionusingimagesegmentationppt-210830184640 (1).pdf
braintumordetectionusingimagesegmentationppt-210830184640 (1).pdf
SunnyYadav735981
 
Brain tumor detection using image segmentation ppt
Brain tumor detection using image segmentation pptBrain tumor detection using image segmentation ppt
Brain tumor detection using image segmentation ppt
Roshini Vijayakumar
 
braintumordetectionusingimagesegmentationppt-210830184640.pdf
braintumordetectionusingimagesegmentationppt-210830184640.pdfbraintumordetectionusingimagesegmentationppt-210830184640.pdf
braintumordetectionusingimagesegmentationppt-210830184640.pdf
AlHussieniAbdulAziz
 
braintumordetectionusingimagesegmentationppt-210830184640.pdf
braintumordetectionusingimagesegmentationppt-210830184640.pdfbraintumordetectionusingimagesegmentationppt-210830184640.pdf
braintumordetectionusingimagesegmentationppt-210830184640.pdf
AlHussieniAbdulAziz
 
Introduction to Machine Learning and Texture Analysis for Lesion Characteriza...
Introduction to Machine Learning and Texture Analysis for Lesion Characteriza...Introduction to Machine Learning and Texture Analysis for Lesion Characteriza...
Introduction to Machine Learning and Texture Analysis for Lesion Characteriza...
Kevin Mader
 
Parkinson disease classification v2.0
Parkinson disease classification v2.0Parkinson disease classification v2.0
Parkinson disease classification v2.0
Nikhil Shrivastava, MS, SAFe PMPO
 
Anits dip
Anits dipAnits dip
Share and Reuse: how data sharing can take your research to the next level
Share and Reuse: how data sharing can take your research to the next levelShare and Reuse: how data sharing can take your research to the next level
Share and Reuse: how data sharing can take your research to the next level
Krzysztof Gorgolewski
 
Meta-Analysis -- Introduction.pptx
Meta-Analysis -- Introduction.pptxMeta-Analysis -- Introduction.pptx
Meta-Analysis -- Introduction.pptx
ACSRM
 
Basic image analysis(processing and classification) and visualization using m...
Basic image analysis(processing and classification) and visualization using m...Basic image analysis(processing and classification) and visualization using m...
Basic image analysis(processing and classification) and visualization using m...
Vishwas N
 
ppt.pdf
ppt.pdfppt.pdf
Spatial and Temporal Features of Noise in fMRI
Spatial and Temporal Features of Noise in fMRISpatial and Temporal Features of Noise in fMRI
Spatial and Temporal Features of Noise in fMRI
Vanessa S
 
3D visualisation of medical images
3D visualisation of medical images3D visualisation of medical images
3D visualisation of medical images
Shashank
 

Similar to CSU_comp (20)

Automatic System for Detection and Classification of Brain Tumors
Automatic System for Detection and Classification of Brain TumorsAutomatic System for Detection and Classification of Brain Tumors
Automatic System for Detection and Classification of Brain Tumors
 
Diagnosis Support by Machine Learning Using Posturography Data
Diagnosis Support by Machine Learning Using Posturography DataDiagnosis Support by Machine Learning Using Posturography Data
Diagnosis Support by Machine Learning Using Posturography Data
 
Machine Learning and Deep Contemplation of Data
Machine Learning and Deep Contemplation of DataMachine Learning and Deep Contemplation of Data
Machine Learning and Deep Contemplation of Data
 
Artificial Intelligence in pathology
Artificial Intelligence in pathologyArtificial Intelligence in pathology
Artificial Intelligence in pathology
 
MRI Brain Tumour Classification Using SURF and SIFT Features
MRI Brain Tumour Classification Using SURF and SIFT FeaturesMRI Brain Tumour Classification Using SURF and SIFT Features
MRI Brain Tumour Classification Using SURF and SIFT Features
 
A Review on the Brain Tumor Detection and Segmentation Techniques
A Review on the Brain Tumor Detection and Segmentation TechniquesA Review on the Brain Tumor Detection and Segmentation Techniques
A Review on the Brain Tumor Detection and Segmentation Techniques
 
braintumordetectionusingimagesegmentationppt-210830184640 (1).pdf
braintumordetectionusingimagesegmentationppt-210830184640 (1).pdfbraintumordetectionusingimagesegmentationppt-210830184640 (1).pdf
braintumordetectionusingimagesegmentationppt-210830184640 (1).pdf
 
Brain tumor detection using image segmentation ppt
Brain tumor detection using image segmentation pptBrain tumor detection using image segmentation ppt
Brain tumor detection using image segmentation ppt
 
braintumordetectionusingimagesegmentationppt-210830184640.pdf
braintumordetectionusingimagesegmentationppt-210830184640.pdfbraintumordetectionusingimagesegmentationppt-210830184640.pdf
braintumordetectionusingimagesegmentationppt-210830184640.pdf
 
braintumordetectionusingimagesegmentationppt-210830184640.pdf
braintumordetectionusingimagesegmentationppt-210830184640.pdfbraintumordetectionusingimagesegmentationppt-210830184640.pdf
braintumordetectionusingimagesegmentationppt-210830184640.pdf
 
Introduction to Machine Learning and Texture Analysis for Lesion Characteriza...
Introduction to Machine Learning and Texture Analysis for Lesion Characteriza...Introduction to Machine Learning and Texture Analysis for Lesion Characteriza...
Introduction to Machine Learning and Texture Analysis for Lesion Characteriza...
 
Model
ModelModel
Model
 
Parkinson disease classification v2.0
Parkinson disease classification v2.0Parkinson disease classification v2.0
Parkinson disease classification v2.0
 
Anits dip
Anits dipAnits dip
Anits dip
 
Share and Reuse: how data sharing can take your research to the next level
Share and Reuse: how data sharing can take your research to the next levelShare and Reuse: how data sharing can take your research to the next level
Share and Reuse: how data sharing can take your research to the next level
 
Meta-Analysis -- Introduction.pptx
Meta-Analysis -- Introduction.pptxMeta-Analysis -- Introduction.pptx
Meta-Analysis -- Introduction.pptx
 
Basic image analysis(processing and classification) and visualization using m...
Basic image analysis(processing and classification) and visualization using m...Basic image analysis(processing and classification) and visualization using m...
Basic image analysis(processing and classification) and visualization using m...
 
ppt.pdf
ppt.pdfppt.pdf
ppt.pdf
 
Spatial and Temporal Features of Noise in fMRI
Spatial and Temporal Features of Noise in fMRISpatial and Temporal Features of Noise in fMRI
Spatial and Temporal Features of Noise in fMRI
 
3D visualisation of medical images
3D visualisation of medical images3D visualisation of medical images
3D visualisation of medical images
 

CSU_comp

  • 1. Computational Classification Techniques for Neuroimaging A Machine Learning Based Approach Adrian Smith – Undergraduate Computer Science Department Sonoma State University
  • 2. Fundamentals • Understanding the human brain has been a central theme of human history • By growing our understanding of the brain, we improve our ability to treat diseases (Gur2002) • Understanding the brain helps us be aware of it’s limitations Artist’s Depiction of Neurons UCI Research Courtesy of OSA Student Chapter at UCI Art in Science Contest. Photo by: Ardy Rahman
  • 3. fMRI Scanning • Functional Magnetic Resonance Imaging (fMRI) allows us to measure localized brain activity • This allows one to find relationships between cognition and brain activity • Blood oxygen is used as a measure of activity (BOLD imaging) • This technique produces rich data, but contains high levels of noise CSRB (Keck MRI Center)
  • 4. Data Collection • One major advantage of researching fMRI data is it’s availability on a variety of online locations • We worked with 1452 total brain scans each corresponding to one of 9 categories • The categories refer to the image a subject was observing
  • 5. Analysis Goals • Our goal was to be able to, given the fMRI scan of a subject, predict what image they were observing • This means differentiating scans based on the image the subject is observing • What is the relationship? Haxby2001 Stimulus Images
  • 6. Machine Learning Techniques • Machine learning is an information processing technique • The field of machine learning is at the heart of understanding “Big Data” • We aimed to use modern machine learning techniques to help classify fMRI data.
  • 7. How does Machine Learning Work? • Machine Learning classification focuses on designing algorithms which are trained to categorize objects • This is done by combining some defining characteristics and a label • The algorithm trains on one set of data, and then is tested to see how accurately it can predict the label of some piece of data. • What is the data? By Antti Ajanki AnAj (Own work) [GFDL (http://www.gnu.org/copyleft/fdl.html), CC-BY-SA- 3.0 (http://creativecommons.org/licenses/by-sa/3.0/)
  • 8. Which is active before processing? Unprocessed Active Unprocessed Rest
  • 9. Which is active after processing? Processed Active Processed Rest
  • 10. Preprocessing • We applied masks that came with the dataset in order to focus on the Ventral Temporal cortex, our region of interest • We then applied a polynomial detrender, which eliminates systematic trends, such as signal increase as the machine warms up • This was followed by a key step, z- scoring against the rest position Graph of Normal Distribution Public Domain
  • 11. Classification • We now had to decide how to process the image data • This meant choosing features that best represented the data we sought • We also tested a variety of classification algorithms which would label images based on the chosen feature
  • 12. Features • We started with the our preprocessed values, and then looked at a variety of transforms • We chose the full vector and the PCA reduced version as our main features of interest • Principle Component Analysis (PCA) is a tool to reduce the dimensionality of a dataset PCA Full Vector (Samples) 50 Highest Values Histogram [0.5, .01, -.02, 1.5, 2.0, … -3.0]576 One Volume
  • 13. Experimental Design • Data was split evenly and randomly into training and test • We used several feature vectors to test each classifier • We primarily focused on k Nearest Neighbor (kNN) and Support Vector Machine (SVM) classifiers • Tests were repeated 15 times and scores averaged Train Feature Training Label Trained Classifier Testing Feature Predicted Label Testing Label Comparison Accuracy and Confusion Matrix Classifier
  • 14. kNN vs. SVM • SVM preforms better than kNN • Increase in accuracy is likely due to the weakness of kNN when dealing with high dimensionality SVM on samples, 90.9% accuracy kNN on samples, 75.6% accuracy
  • 15. • We applied PCA to the processed data • This produced a vector over half the size of our original • This smaller vector produces more accurate results Samples vs. PCA PCA (SVM), 92.1% accuracy SVM on samples, 90.9% accuracy
  • 16. • PCA and SVM in combination gave the best results after repeated testing • We achieved on average 92.1% accuracy among 9 labels, with a 2.0% standard deviation. • Our classification methods are effective and repeatable • We also gained a variety of insights about the nature of the data Classification Results: Accuracy
  • 17. • We saw several labels which repeatedly misclassified, and saw accuracy improve as they were removed • One area of further study is investigating whether these patterns exist between multiple subjects, and why PCA (SVM), 92.1% accuracy Classification Results: Insights
  • 18. Future Exploration • We intend to move towards classifying across multiple subjects • This is of utmost importance to clinical applications of fMRI data • Multisubject comparison presents challenges due to the variation in brain structure • We intend to build upon previous work on feature detection and scaling maps (Gill2014)
  • 19. Sources • Gur, R. E., McGrath, C., Chan, R. M., Schroeder, L., Turner, T., Turetsky, B. I., ... & Gur, R. C. (2002). An fMRI study of facial emotion processing in patients with schizophrenia. American Journal of Psychiatry. • Haxby, J. V., Gobbini, M. I., Furey, M. L., Ishai, A., Schouten, J. L., & Pietrini, P. (2001). Distributed and overlapping representations of faces and objects in ventral temporal cortex. Science, 293(5539), 2425-2430. • Gill, G., Bauer, C., & Beichel, R. R. (2014). A method for avoiding overlap of left and right lungs in shape model guided segmentation of lungs in CT volumes. Medical physics, 41(10), 101908. • Dataset: This data was obtained from the OpenfMRI database. Its accession number is ds000105. The original authors of :ref:`Haxby et al. (2001) <HGF+01>` hold the copyright of this dataset and made it available under the terms of the `Creative Commons Attribution-Share Alike 3.0`_ license.
  • 20. Acknowledgments • Dr. Gurman Gill – Mentor • OpenfMRI – Source of all data, and amazing example of open data in science • pyMVPA – Python toolkit used in preprocessing • Scikit-learn – Python toolkit used in classification • Dr. Yaroslav Halchenko – Researcher who provided extensive aid in understanding and dealing with fMRI data
  • 22. Extra Graphics SVM of top 400 values. 30.9% accuracy SVM on 90% PCA. 92.2% accuracy