Learning based image segmentation of post-operative CT-images: A hydrocephalus case study

Accurate estimation of volumes for cerebrospinal fluid (CSF) and brain before and after surgery (pre-op and post-op) plays an important role in analyzing treatment for hydrocephalus. This in turn, relies upon segmentation of brain imagery into brain tissue and CSF. Segmentation of preop images is a relatively straightforward problem and has been well researched. However, segmenting post-op CT-scans becomes challenging due to distorted anatomy and subdural hematoma collections pressing on the brain. Most intensity and feature based segmentation methods fail to separate subdurals from brain and CSF as subdural geometry varies greatly across different patients and their intensity varies with time. We combat this problem by a learning approach that treats segmentation as supervised classification at the pixel level, i.e. a training set of CT scans with labeled pixel identities is employed. Inspired by sparsity constrained classification, our central contribution is a dictionary learning framework that learns class (segment) specific dictionaries that can efficiently represent test samples from the same class while poorly represent corresponding samples from other classes. Because discriminating features are discovered automatically, we call our method feature learning for image segmentation (FLIS). Experiments performed on infant CT brain images acquired from CURE children's hospital of Uganda reveal the success of our method against existing alternatives.