Learning Morphological Feature Perturbations for Calibrated Semi-Supervised Segmentation

TL;DR

MisMatch introduces a semi-supervised segmentation method that uses learned feature perturbations and consistency regularization, leading to improved accuracy and calibration on medical imaging tasks.

Contribution

It presents a novel framework with feature perturbations and dual decoders for better semi-supervised segmentation performance.

Findings

Outperforms state-of-the-art methods on pulmonary vessel segmentation

Achieves superior results on brain tumour segmentation

Provides better model calibration than supervised learning

Abstract

We propose MisMatch, a novel consistency-driven semi-supervised segmentation framework which produces predictions that are invariant to learnt feature perturbations. MisMatch consists of an encoder and a two-head decoders. One decoder learns positive attention to the foreground regions of interest (RoI) on unlabelled images thereby generating dilated features. The other decoder learns negative attention to the foreground on the same unlabelled images thereby generating eroded features. We then apply a consistency regularisation on the paired predictions. MisMatch outperforms state-of-the-art semi-supervised methods on a CT-based pulmonary vessel segmentation task and a MRI-based brain tumour segmentation task. In addition, we show that the effectiveness of MisMatch comes from better model calibration than its supervised learning counterpart.