Brain Imaging MESA Science Symposium September 22 – 23, 2011

1. Brain ImagingMESA Science SymposiumSeptember 22 – 23, 2011 R. Nick Bryan, MD, PhD Univ. Pennsylvania Disclosures: Departmental grants: GE, Phillips, Siemens Departmental IP: AVID

2. CerebroVascular Disease • Pathology • Large vessel disease • Transcortical stroke • Immediate, significant clinical deficits • Relatively low prevalence (< 5 %, 65 yo) • Small vessel ischemic disease (SVID) • Lacunar disease • Acute infarct • Lacunar state, multi-infarct dementia • Leukoaerosis • Indolent onset, subclinical and clinical deficits • Relatively high prevalence (> 20 %, 65 yo) • Cognitive decline • MR defined • Major impact on individuals and society

3. MRI and CVD • Non-invasive diagnosis and documentation of CVD • Morphological imaging • Large vessel • Small vessel disease (MR requisite) • Physiological imaging • Potentially new pathophysiological features of subclinicial and clinical disease • Important epidemiological and clinical trial tool (CHS, ARIC, Framingham, AGES, Rotterdam, ACCORD, SPRINT, ADNI) • Natural history disease • Risk factors • Correlates • Clinical and Subclinical disease • Cognitive impairment/dementia • Major event predictor • Future stroke • Cognitive decline

4. FLAIR T2 Raw Images Derived Images 3 T Brain MRIProtocolVariables T1 DTI GM, WM, CSF Ischemic Lesions Morphology Anatomy/Gross Path BOLD-BH BOLD-Rest ASL Morphology WM Integrity FA TR Physiology Blood Flow CBF Physiology Vascular Reactivity % Signal Change Physiology Connectivity Functional Connectivity

5. CHS/ARIC White Matter Grades

6. Tissue/Lesion Classification • Automated calculation of the probability of a voxel belonging to anormal tissue or lesion class.b 1.0 0.5 Thresholding 0.0 Probability map overlaid on FLAIR image Thresholded probability map overlaid on FLAIR image

7. Image Processing Pipeline ANALYSIS/SEGMENTATION Standard Atlas (partitioned to 106 ROI’s) Segment- ation Skull Stripping Atlas-based Labeling Labeled image of the individual Statistical analysis of ROI volumes and other properties (e.g. signal characteristics)

8. Brain MRI Variables

9. CARDIA Brain MRI Variables • Morphometric analysis • 92 Anatomic Regions of Interest (ROI’s) • CSF • Basal Ganglia • Normal • Abnormal (Lacunar infarcts, SVID) • Necrotic/non-necrotic • White Matter • Normal • Abnormal (Infarcts, SVID) • Necrotic/non-necrotic • Gray Matter • Normal • Abnormal (Infarcts, SVID) • Necrotic/non-necrotic • Additional variables • Morphological • DTI – FA, ADC (scalar) • Physiological • ASL – CBF (scalar) • fMRI – vascular reactivity, regional connectivity (scalar, vector) • Options • Voxel based analysis

10. Brain MRI Physiological Parameters • Most unique MRI data • Not previously performed in major epidemiological studies • CHS, ARIC, MESA, Framingham, Rotterdam • Added to year 25 CARDIA • Cerebral Blood Flow • Arterial spin label technique • Vascular Reactivity • Breath Hold fMRI • Functional Connectivitiy • Resting fMRI

11. Cerebral Blood FlowArterial Spin Label (ASL) • CBF Model: Kety/Schmidt • Diffusible tracer technique • H2O15 PET, Xe CT • MRI Arterial spin label H2O • Detre et al • CARDIA pCASL Protocol • Wang et al • Pulse sequence and data analysis

12. Arterial Spin Labeled Perfusion MRI }1% effect

13. Cerebrovascular Reactivity • Cerebrovascular autoregulation • Maintains constant CBF in steady state • Adjusts to meet local brain activity • CO2 primary control • Functional MRI (fMRI) • Blood Oxygen Level Dependent (BOLD) fMRI • Signal intensity dependent on CBF and O2 concentration • Box car task experimental design • Neural task • Breath Hold task

14. Vascular Reactivity:Breath Hold fMRI • BOLD box car fMRI • Hypothesis: BH > increased CO2 > increased CBF & decreased O2 > increased BOLD signal • Task: 30 second normal respiration/16 second Breath Hold, 4 repeats. • Regional and global change • BH fMRI analysis • SPM ROI analysis • Application of general linear model to change in signal over time • Superior Sagittal Sinus analysis for global response

15. BH modulates cerebral blood flow to all vascularized brain regions without an accompanying change in CMRO2 BH with BOLD can be used for accessing global and/or regional vascular reactivity properties SSS BOLD signal reflects global response BH fMRISuperior Sagittal Sinus (SSS)

16. CARDIA Participant BH ResponsesN = 50

17. Brain Functional Connectivity • Cerebral functional connectivity inferred by temporally correlated brain measures • Microelectrodes, EEG, MEG • BOLD fMRI • Resting fMRI studies investigate functional connectivity between spatially distinct but functionally related regions • Comparing correlation coefficients of pairs of regions • Low frequency fluctuations(LFF) of BOLD signal in specific brain areas • Normal subjects have • Positive correlations between the prefrontal and parietal lobes on resting fMRI • Default Mode • Negative correlations within the prefrontal, parietal lobes and occipital lobes. • Compared with healthy controls, AD patients have • Decreased positive correlations between the prefrontal and parietal lobes on fMRI • Increased positive correlations within the prefrontal, parietal lobes and occipital lobes. • Decreased negative correlation between two intrinsically anti correlated networks.

18. Resting fMRI DM in AD Control Mild AD Mod AD Severe AD Zhang, 2010

19. PET Brain Imaging • Physiology • CBF – H215O • Metabolism • Glucose – 18FDG • Molecular markers • B amyloid • 11C: PiB: 20 minute half life • 18F: [F-18]BAY94-9172 , [F-18]AV-45, [F-18]AH110690 (or [F-18]3′F-PiB): 90 minute half life • Neurodegenerative disease (AD) • Vs CVD • Amyloid angiopathy

20. MESA Rationale & Objectives • Study of atherosclerosis • Risk factors, predictors • Subclinical and clinical disease • Heart disease, stroke and other diseases of blood vessels • New information about pathophysiology of subclinical disease development and progression and its role in clinical cardiovascular disease • Suggest new interventions to prevent progression of subclinical disease and conversion to clinical disease

21. MESA Objectives (select) • Primary • Determine characteristics related to progression of subclinical to clinical cardiovascular disease • Secondary • Assess ethnic, age, gender differences in subclinical disease prevalence and risk progression and clinical cardiovascular disease • Describe interrelationships of newly identified and established risk factors and measures of subclinical disease • Develop population-based methods for future screening and intervention studies.

22. MESA Protocol • 6,000 men and women from 6 geographic regions, aged 55 to 95 • Diverse race and ethnic communities • H and P • Std coronary risk, sociodemographic, life style, psychosocial factors, life style factors • Coronary ca (CT), ventric mass/function, (MR), carotid artery disease (MR, US), vascular insufficiency (fm brachial vasodilation, A/B BP, arterial wave form), ECG • Blood samples, DNA • Outcome events: CHD/MI/CHF, stroke, mortality • Visit 5: Neuropsychological testing, visual assessment and retinal photography

23. MESA Brain Imaging Scientific Value • Morphological MRI • Novel quantitative measures of large and small vessel CVD plus white matter integrity/connectivity variables, FA/ADC. • Refined morphological descriptive and correlative analysis • Physiological MRI data, CBF/VR/RfMRI. • Novel physiological parameters allow investigation of pathophysiological precedents of morphological disease • Possibly more sensitive or earlier markers of CVD and its functional sequela • Important population based B amyloid data • CVD and neurodegenerative disease

25. CARDIA/SPRINT 3T Brain MRI Protocol