qualifying exam presentation

37
Data Driven Neuropsychiatric Profiling Qualifying Exam Presentation Vanessa Sochat August 19, 2013

Upload: vanessa-sochat

Post on 13-Apr-2017

278 views

Category:

Science


1 download

TRANSCRIPT

Page 1: Qualifying Exam Presentation

Data Driven Neuropsychiatric Profiling

Qualifying Exam PresentationVanessa SochatAugust 19, 2013

Page 2: Qualifying Exam Presentation

Why do we need a model of disorder?Treatment hinges on diagnostic subtypes

patient assessment

autismADHDbipolar

schizophrenia

diagnosis of

bipolar

treatment

education

finances

research

?diagnosi

s of bipolar

treatment

education

finances

research

Page 3: Qualifying Exam Presentation

The Diagnostic and Statistical Manual of Mental Disorder (DSM)

“loss of interest”“depressed mood”“insomnia”“excessive sleeping”

depression

anxiety

PTSDdepressio

n

autismADHD

bipolar

Page 4: Qualifying Exam Presentation

Autism Spectrum Disorder (ASD):A childhood development disorder

• Afflicts 1 in 100 children • Economic burden of $126 billion annually• Social, communication, and cognitive deficits,

repetitive behaviors and interests

data-driven subtyping of autism spectrum disorders for early diagnosis and tailored, effective treatmentAutism

Asperger’s

PDD-NOS

DSM IV

Autism

DSM V

Page 5: Qualifying Exam Presentation

Unsolved Problem:

data-driven subtyping of neuropsychiatric disorder for early diagnosis and tailored, effective treatment

Page 6: Qualifying Exam Presentation

How might we model disorder?

Genetics

Behavior

Neuroimaging

combinations of genes small effects high heritability a marker of risk, not certainty

biased and unreliable

brain phenotype to define disorder?

Page 7: Qualifying Exam Presentation

Goal

A new model of neuropsychiatric disorder based on patterns of local brain structure

neuropsychiatric profile

brain phenotyp

e

cognitive phenotyp

e

Page 8: Qualifying Exam Presentation

Hypothesis

Mining imaging data will define clinically-useful disease subtypes better than currently possible

using DSM alone.

Page 9: Qualifying Exam Presentation

Aim1: to develop a computational representation of brain phenotype, and use it to define subtypes of neurological disorder

Aim 2: to create a structured normalized representation of cognitive phenotype as a new standard for evaluationAim 3: to evaluate by reproducing known brain-behavioral relationships, and finding subtypes of autism

Specific Aims

Page 10: Qualifying Exam Presentation

Aim1: to develop a computational representation of brain phenotype, and use it to define subtypes of neurological disorder

brain phenotyp

e

cognitive phenotyp

e

Page 11: Qualifying Exam Presentation

DISORDER HEALTHYCONTROL

Start with groupsCollect dataFind differencesInconsistent results

How do we find brain-based differences?

1234

Page 12: Qualifying Exam Presentation

Assumptions about the brain

Assumptions about the groups

How do we find brain-based differences?

make assumptions about spatial meaningfulness voxel-wise features prime to overfitting

diagnostic categories represent homogenous syndromes what about heterogeneity of the healthy controls?

learn groups from the data

make no assumptions about spatial location strive for an abstraction of structure

Page 13: Qualifying Exam Presentation

Collect dataLocal brain phenotypeDefine groupsFind Differences

MRI

How should we find these differences?

1234

1

GROUP A GROUP B4

2

3

groups == proposed subtypes

Page 14: Qualifying Exam Presentation

Aim 2: to create a structured normalized representation of cognitive phenotype as a new standard for evaluation

brain phenotyp

e

cognitive phenotyp

e

Page 15: Qualifying Exam Presentation

BEHAVIORAL & COGNITIVE METRICS

Structure and normalize dataLink to specific traitsCreate cognitive phenotypes

Assessment metrics reflect human traits

123

Assessment 1

Q1…Q10 “Anxiety”

Assessment 2

Q30…Q40 “Anxiety”

2

IQ

anxiety, attention …

T1 T2 T3 …. TN

1

3cognitive phenotyp

e

Page 16: Qualifying Exam Presentation

Aim 3: to define and evaluate subtypes of autism spectrum disorderAim 3: to evaluate by reproducing known brain-behavioral relationships, and finding subtypes of autism

brain phenotyp

e

cognitive phenotyp

e

Page 17: Qualifying Exam Presentation

C1 C2 C3 …. CN

Evaluation of Brain Phenotype:Reproduce known Brain-Behavior Relationships

C1 C2 C3 …. CNT1 T2 T3 …. TN

0.20.50.7

2.9 0.20.1 1.97.2 2.1

Person APerson BPerson C

X1 X2 Y

Page 18: Qualifying Exam Presentation

ClusteringASD

CONTROL

Evaluation of Method:Show that we capture current gold standard

Page 19: Qualifying Exam Presentation

my groupsDSM

Evaluation of Method:Show greater homogeneity in our groups

different

similar

Page 20: Qualifying Exam Presentation

ASD CONTROL

GROUP A GROUP B

my groups

DSM

Evaluation of Method:Show more robust group differences

significant differences?

vs

vs

Page 21: Qualifying Exam Presentation

Summary:Data-driven neuropsychiatric profiling

patient

brain phenotype

MRI diagnosis

treatment

education

finances

research

Page 22: Qualifying Exam Presentation

Future Vision:Extension to other disorders and data types

brain phenotyp

e

AutismWhite MatterADHDBipolar

Gray MatterGene Expression

EHRdoctor visitsi-phone app

geneticmolecularclinical

cognitive phenotyp

e

Page 23: Qualifying Exam Presentation

Informatics Contribution

• Tools to extend Big Data paradigms to neuroscience

• Data-driven model of neuropsychatric profile• Brain phenotype• Cognitive phenotype

• Methods to define subtypes of disorder

Page 24: Qualifying Exam Presentation

Biological Contributions

• Discovery of biomarkers of disorder• Definition of disorder subtypes• Enabling personalized treatment

Page 25: Qualifying Exam Presentation

Acknowledgements

Advisors and PanelDaniel RubinRuss AltmanMark MusenAntonio Hardan

ColleaguesKaustubh SupekarFeature GroupThe MIND Institute

Support StaffJohn DiMarioMary Jeanne & Nancy

FundingMicrosoft ResearchSGF and NSF

Friends and Fellow BMIFrancisco GimenezRebecca SawyerLuke Yancy Jr.Beth PerchaKatie PlaneyTiffany Ting LuLinda SzaboDiego MunozJonathan MortensenThe M&Ms previously known as first years

Page 26: Qualifying Exam Presentation

Thank you!

Page 27: Qualifying Exam Presentation

2

anxiety, attention …

C1 C2 C3 …. CNAssessment X

Subscale Y “Anxiety”

Assessment A

Subscale B “Anxiety”

Cognitive Phenotypes for N = 26

Page 28: Qualifying Exam Presentation

…are meaningful about groups

Page 29: Qualifying Exam Presentation

(OBSERVED DATA)

(MIXING MATRIX) (ORIGINAL DATA)

X = A SX

n x m n x n n x m

fMRI data

time

time

space spacecomponents

components

spatial maps

How do we discover local brain structure?Matrix Decomposition

Page 30: Qualifying Exam Presentation

n x m n x n n x m

How do we discover local brain structure?Matrix Decomposition

sMRI data

brains

brains

components

components

space space

spatial maps

2.9 0.2 1.9brain

phenotype

Page 31: Qualifying Exam Presentation

Clustering GROUP 1

GROUP 2

How do we define subtypes?

GROUP 3GROUP 4

Page 32: Qualifying Exam Presentation

BEHAVIORAL & COGNITIVE METRICS

1. Collect data2. Covariate or control3. Data not used

How do we use assessment metrics?

123

Page 33: Qualifying Exam Presentation

A Mapping from Assessments to TraitsMapping Assessments to TraitsThe Cognitive Atlas

Page 34: Qualifying Exam Presentation

Mapping Assessments to TraitsThe Cognitive Atlas

Page 35: Qualifying Exam Presentation

Mapping Assessments to TraitsThe Cognitive Atlas

Page 36: Qualifying Exam Presentation

Representation Based on TraitsThe Cognitive Phenotype

1 32

Assessment X Subscale X “Anxiety” .04 Subscale Y “Attention” .90

Normalized Score =Total - Raw

Total

Page 37: Qualifying Exam Presentation

3

Representation Based on TraitsThe Cognitive Phenotype

C1 C2 C3 …. CN

Find all assessments for person LOUIE…

4

Find me a cohort defined by anxiety > .5

Are anxious people also impulsive?

brain phenotyp

e