Decoupling Shape and Density for Liver Lesion Synthesis Using Conditional Generative Adversarial Networks

TL;DR

This paper introduces a novel method for liver lesion synthesis that decouples shape and density, enabling better control over generated samples and improving segmentation performance.

Contribution

It proposes a framework that separates shape and density in lesion synthesis using conditional GANs, enhancing control and diversity of generated data.

Findings

Decoupling shape and density allows for targeted lesion synthesis.

Embedding density information improves segmentation accuracy.

Qualitative results show effective control over lesion features.

Abstract

Lesion synthesis received much attention with the rise of efficient generative models for augmenting training data, drawing lesion evolution scenarios, or aiding expert training. The quality and diversity of synthesized data are highly dependent on the annotated data used to train the models, which not rarely struggle to derive very different yet realistic samples from the training ones. That adds an inherent bias to lesion segmentation algorithms and limits synthesizing lesion evolution scenarios efficiently. This paper presents a method for decoupling shape and density for liver lesion synthesis, creating a framework that allows straight-forwardly driving the synthesis. We offer qualitative results that show the synthesis control by modifying shape and density individually, and quantitative results that demonstrate that embedding the density information in the generator model helps to increase lesion segmentation performance compared to using the shape solely.