Prior-based Coregistration and Cosegmentation

TL;DR

This paper introduces a flexible framework for dense coregistration and cosegmentation that leverages classifier outputs and population registration to improve alignment and segmentation of brain structures.

Contribution

It presents a modular approach that replaces ground truth with classifier outputs and combines them with population deformable registration, enhancing segmentation accuracy.

Findings

Effective on challenging brain datasets

Improves alignment and segmentation quality

Compatible with various classifiers and metrics

Abstract

We propose a modular and scalable framework for dense coregistration and cosegmentation with two key characteristics: first, we substitute ground truth data with the semantic map output of a classifier; second, we combine this output with population deformable registration to improve both alignment and segmentation. Our approach deforms all volumes towards consensus, taking into account image similarities and label consistency. Our pipeline can incorporate any classifier and similarity metric. Results on two datasets, containing annotations of challenging brain structures, demonstrate the potential of our method.