SA-GAN: Structure-Aware GAN for Organ-Preserving Synthetic CT Generation

TL;DR

SA-GAN is a novel deep learning model that generates synthetic CT images from MRI while preserving organ shapes and locations, aiding MR-only treatment planning despite internal tissue changes.

Contribution

The paper introduces SA-GAN, a structure-aware GAN that maintains organ integrity during MRI-to-CT synthesis, addressing inconsistencies caused by tissue changes.

Findings

SA-GAN achieves clinically acceptable accuracy in synthetic CT generation.

SA-GAN effectively preserves organ shapes and locations.

Supports MR-only treatment planning in variable internal organ conditions.

Abstract

In medical image synthesis, model training could be challenging due to the inconsistencies between images of different modalities even with the same patient, typically caused by internal status/tissue changes as different modalities are usually obtained at a different time. This paper proposes a novel deep learning method, Structure-aware Generative Adversarial Network (SA-GAN), that preserves the shapes and locations of in-consistent structures when generating medical images. SA-GAN is employed to generate synthetic computed tomography (synCT) images from magnetic resonance imaging (MRI) with two parallel streams: the global stream translates the input from the MRI to the CT domain while the local stream automatically segments the inconsistent organs, maintains their locations and shapes in MRI, and translates the organ intensities to CT. Through extensive experiments on a pelvic dataset, we demonstrate that SA-GAN provides clinically acceptable accuracy on both synCTs and organ segmentation and supports MR-only treatment planning in disease sites with internal organ status changes.