120 likes | 278 Views
Statistical Models of Anatomy and Pathology. Polina Golland. Statistical Models of Anatomy. Applications Spatial priors for segmentation Population studies Traditional approach Align images to a common template Compute mean and co-variation Challenges
E N D
Statistical Models of Anatomy and Pathology Polina Golland
Statistical Models of Anatomy • Applications • Spatial priors for segmentation • Population studies • Traditional approach • Align images to a common template • Compute mean and co-variation • Challenges • Spatial variability in the structure of interest • Loss of detail • Heterogeneous populations
Our Solutions • Use training data in novel ways • handle spatial variability • TBI, tumors • avoid the loss of detail • Atrial Fibrillation, Huntington’s, Alzheimer’s • Model heterogeneous populations • capture broader variability • Atrial fibrillation, radiation therapy, Alzheimer’s • Related topics • Registration (Guido Gerig) • Interactive segmentation (Allen Tannenbaum)
Spatial Priors and Pathology • Augmented generative model • Atlas: spatial prior for healthy tissues • Estimate: spatial prior for tumor • Output • Common healthy tissue segmentation • Modality-specific tumor segmentation Menze, MICCAI 2010
Spatial Priors and Pathology (cont’d) • More accurate than EM-segmentation with outlier detection • Comparable to within-rater variability • Going forward: TBI Menze, MICCAI 2010
Label Fusion Segmentation Test Image Pairwise Registration Subject Specific Label Prior New Segmentation Training Data
… … Training images M {Ln} {In} Test image ? L(x) I(x) Generative Model for Label Fusion Sabuncu, TMI 2010
Left Atrium Segmentation Manual Parametric Majority Weighted fusion • More accurate than baseline methods • Correctly identified all veins • Local prior for scar location Depa, MICCAI Workshop 2010
Modeling Heterogeneous Populations • Manifold of anatomical images • Spectral embedding • Statistical model in new space • Gerber, MedIA 2010 noise • Collection of sub-populations • Mixture model • Templates represent population • Sabuncu TMI 2009
Applications for Spatial Priors • Identify relevant “neighborhood” for the new image • A (small) set of training examples • A (local) atlas template • Construct patient-specific spatial prior • Average or use label fusion • Challenges: • Reduce the number of pairwise registration steps • Model influence of selected neighborhood on new image
Conclusions • Clear need for new methods • Handle spatial variability of pathology • Handle anatomical variability in a population • Preliminary results: local models • In the image coordinates • In the space of images • Going forward • Development in the context of the DBPs