integrative multi-scale analysis in biomedical data science: tools, methods and challenges
TRANSCRIPT
Integra(ve Mul(-‐scale Analysis in Biomedical Data Science:
Tools, Methods and Challenges
Joel Saltz Department of Biomedical Informa(cs
Stony Brook University CI4CC October 2015
Targeted Therapy against bcr-‐abl -‐-‐ Leukemia (CML)
Intertumor and intratumor heterogenity
Burrell et al. Nature (2013):338–345
Bruin et al, Swanton Science 2014
Slide – thanks to Adam Marcus
MulI-‐scale IntegraIve Analysis in Precision Medicine
• Predict treatment outcome, select, monitor treatments
• Reduce inter-‐observer variability in diagnosis
• Computer assisted exploraIon of new classificaIon schemes
• MulI-‐scale cancer simulaIons
From Daniel Rubin’s Website
Example: Pathology AnalyIcal Imaging
• Provide rich informaIon about morphological and funcIonal characterisIcs
• Image analysis, feature extracIon on mulIple scales • SpaIally mapped “omics” • MulIple microscopy modaliIes
Glass Slides Scanning Whole Slide Images Image Analysis
Integra<ve Morphology/”omics”
Quantitative Feature Analysis in Pathology: Emory In Silico Center for Brain Tumor Research (PI = Dan Brat, PD= Joel Saltz) NLM/NCI: Integrative Analysis/Digital Pathology R01LM011119, R01LM009239 (Dual PIs Joel Saltz, David Foran) Marcus Foundation Grant – Ari Kaufman, Joel Saltz
Associations
Radiomics
Decoding tumour phenotype by noninvasive imaging using a quan<ta<ve radiomics approach
Hugo J. W. L. Aerts et. Al. Nature Communica/ons 5, ArIcle number: 4006 doi:10.1038/ncomms5006
Features
Pa<ents
The Driving Meta Applica(on
IdenIfy and segment trillions of mulI-‐scale objects from spaIo-‐temporal datasets Extract features from objects and spaIo-‐temporal regions Support queries against ensembles of features extracted from mulIple datasets StaIsIcal analyses and machine learning to link features to physical and biological phenomena Feature driven simulaIon – use extracted features as simulaIon iniIal, boundary condiIons and to assimilate data into simulaIons
Detect and track changes in data during production
Invert data for reservoir properties Detect and track reservoir changes
Assimilate data & reservoir properties into
the evolving reservoir model Use simulation and optimization to guide future production
Example: Oil Field Management – Joint NSF ITR with Mary Wheeler, Paul Stoffa
Commonali(es with Physical Science and Engineering
• MulI-‐scale material/Issue structural, molecular, funcIonal characterizaIon. Design of materials with specific structural, energy storage properIes, brain, regeneraIve medicine, cancer
• IntegraIve mulI-‐scale analyses of the earth, oceans, atmosphere, ciIes, vegetaIon etc – cameras and sensors on satellites, aircra^, drones, land vehicles, staIonary cameras
• Digital astronomy • Hydrocarbon exploraIon, exploitaIon, polluIon remediaIon
• Aerospace – wind tunnels, acquisiIon of data during flight
• Solid prinIng integraIve data analyses • Autonomous vehicles, e.g. self driving cars • Data generated by numerical simulaIon codes – PDEs, parIcle methods
• Mul$-‐scale Precision Medicine
Thanks!