Disentangled PET Lesion Segmentation

TL;DR

This paper introduces PET-Disentangler, a 3D neural network that improves PET lesion segmentation by disentangling healthy and diseased features, reducing false positives and enhancing accuracy in identifying cancerous lesions.

Contribution

The novel PET-Disentangler method employs a 3D disentanglement approach with a critic network to better distinguish healthy tissue from lesions in PET images.

Findings

Reduces false positives in lesion detection.

Improves segmentation accuracy over existing methods.

Effectively disentangles healthy and diseased features.

Abstract

PET imaging is an invaluable tool in clinical settings as it captures the functional activity of both healthy anatomy and cancerous lesions. Developing automatic lesion segmentation methods for PET images is crucial since manual lesion segmentation is laborious and prone to inter- and intra-observer variability. We propose PET-Disentangler, a 3D disentanglement method that uses a 3D UNet-like encoder-decoder architecture to disentangle disease and normal healthy anatomical features with losses for segmentation, reconstruction, and healthy component plausibility. A critic network is used to encourage the healthy latent features to match the distribution of healthy samples and thus encourages these features to not contain any lesion-related features. Our quantitative results show that PET-Disentangler is less prone to incorrectly declaring healthy and high tracer uptake regions as cancerous lesions, since such uptake pattern would be assigned to the disentangled healthy component.