Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

NEUROINFORMATICS AND THE COMPLEXITY OF THE BRAIN AT THE MORPHOME (200휇휇) SCALE

14 views

Published on

2016 ORAU Annual Meeting of the Council of Sponsoring Institutions
Michael I. Miller
Kavli Neuroscience Discovery Institute

Published in: Government & Nonprofit
  • Be the first to comment

  • Be the first to like this

NEUROINFORMATICS AND THE COMPLEXITY OF THE BRAIN AT THE MORPHOME (200휇휇) SCALE

  1. 1. NEUROINFORMATICS AND THE COMPLEXITY OF THE BRAIN AT THE MORPHOME (200𝜇𝜇) SCALE 2/14/2005 SAND7 2015 Michael I. Miller Kavli Neuroscience Discovery Institute
  2. 2. Supported by NIH NINDS NIH NIMH NIH NIA NIH NIBIB NSF XSEDE Kavli Foundation
  3. 3. High Throughput Neuroinformatic BrainClouds www.MRICloud.org Mori,Oishi, Faria Miller, Faria, Oishi, Mori, “High Throughput Neuroinformatics”. Frontiers Neuroinformatics, 2013.
  4. 4. HIGH THROUGHPUT NEUROINFORMATICS Atlas Modalities T1,T2,DWI, PET, fMRI GPS POSITIONING SYSTEM
  5. 5. We position information across scales using diffeomorphisms. High Field 11.7T 200𝜇𝜇 Histology 𝜇𝜇DTI
  6. 6. High throughput positioning is a computational fluid-dynamics “like” code. The positioning engine is a Software as a Service hosted at JHU and the NSF XSEDE: eXtreme Science Engineering Discovery Environment.
  7. 7. eye mouth Left Eye Lower Lip Flows Generate Diffeomorphisms Positioning via Diffeomorphisms
  8. 8. Knowledge Ontology Brain Atlases NeuroCloud • Pediatric • Geriatric BrainGPS Positioning System Machine Learning Diseases Brain Cloud 3/21/2016 WWW.MRICLOUD.ORG Miller, Faria, Oishi, Mori, High Throughput Neuroinformatics”. Frontiers Neuroinfromatics, 2013.
  9. 9. T2 FA MD Vol T2 FA MD Vol T2 FA MD Vol T2 FA MD Vol BRAINS BECOME 1000 BYTE INDEX
  10. 10. COLLECT MANY TEXT RECORDS FROM DATA BASE MORI AND MILLER , JOHNS HOPKINS UNIVERSITY
  11. 11. Tissue Atrophy Cerebrum (neo cortex areas) Hemisphere Left -4.00 Right -3.34 frontal lobe L/R Left Type2 Level3' -3.34 Right -3.17 parietal lobe L/R Left Type2 Level3' -2.88 Right -1.87 temporal lobe L/R Left Type2 Level3' 2.67 Right 2.65 occipital lobe L/R Left Type2 Level3' 0.90 Right 1.32 insula L/R Left Type2 Level3' -1.11 Right -0.56 inferior deep white matter L/R Left Type2 Level3' -2.79 Right -1.43 posterior deep white matter L/R Left Type2 Level3' -3.70 Right -2.72 Cerebrum (limbic) L/R limbic (cingulate) L/R Left Type1 Level4 -2.69 Right -2.61 limbic (hippocampus) L/R Left Type1 Level4 -0.16 Right -1.73 limbic (medial temporal) L/R Left Type1 Level5 1.44 Right -0.57 limbic (amygdala) L/R Left Type1 Level4 -1.26 Right -0.23 basal ganglia L/R basal ganglia (caudate nucleus) L/R Left Type1 Level4 2.05 Right 2.24 basal ganglia (putamen) L/R Left Type1 Level4 -2.42 Right -3.33 basal ganglia (globus pallidus) L/R Left Type1 Level4 -1.95 Right -2.01 thalamus L/R thalamus L/R Left Type1 Level3 -2.73 Right -3.47 corpus callosum corpus callosum Left Type1 Level3 -0.40 Right 0.04 cerebellum cerebellum Type1 Level3 -0.12 midbrain midbrain (tectum + tegmentum) Left Type1 Level5 -2.82 Right -2.52 pons pons Left Type1 Level5 -1.81 Right -1.47 Expansion lateral ventricle L/R anterior horn of lateral ventricle L/R Left Type1 Level5 5.15 Right 5.03 body of lateral ventricle L/R Left Type1 Level5 3.95 Right 4.21 posterior horn of lateral ventricle L/R Left Type1 Level5 1.19 Right 2.63 3rd ventricle 3rd ventricle Type1 Level5 2.70 1 million voxels 1000 byte vector Clinical filter Language generation Severe bi-hemispheric atrophy. Severe frontal lobe atrophy. Temporal lobe and occipital lobe are preserved. Severe ventricle expansion in the anterior horn of the lateral ventricles and the body of the lateral ventricles. Cloud says
  12. 12. Neurodevelopmental diseases have brain changes which start many years before clinical symptoms. …aging does as well.
  13. 13. T1 Gray matter: Aging SUPERIOR PARIETAL GYRUS MEDIAL GYRUS PRECENTRAL GYRUS POST CENTRAL GYRUS RECTUS GYRUS Gray matter spaces are decreasing!
  14. 14. These statistics form the basis for Personalized Radiological Health Care.
  15. 15. Click on one roi, a popup menu appears -3 3 MRICloud: Patient Specific
  16. 16. MRICloud: Patient Specific
  17. 17. Machine learning for clustering into heteregeneous neurological diseases.
  18. 18. Alzheimer’s-ProgressiveAphasia-Huntingtons-Ataxia Alzheimer’s disease Huntington’s diseaseProgressive Aphasia Spinocerebellar Ataxia type 6 hippocampus temporal parietal Left Hemi. Language area Basal ganglia cerebellum
  19. 19. our developing children…
  20. 20. Cerebral Palsy and Machine Learning Volume Mean Diffusivity Fractional Anisotropy Multi-Modal Classifier
  21. 21. Thank-you

×