Improving rectal tumor segmentation with anomaly fusion derived from anatomical inpainting: a multicenter study.

Abstract

Accurate rectal tumor segmentation using magnetic resonance imaging (MRI) is paramount for effective treatment planning. It allows for volumetric and other quantitative tumor assessments, potentially aiding in prognostication and treatment response evaluation. Manual delineation of rectal tumors and surrounding structures is time-consuming and labor-intensive. Over the past few years, deep learning has shown strong results in automated tumor segmentation in MRI. Current studies on automated rectal tumor segmentation, however, focus solely on tumoral regions without considering the rectal anatomical entities and often lack a solid multicenter external validation. In this study, we improved rectal tumor segmentation by incorporating anomaly maps derived from anatomical inpainting. This inpainting was trained using a U-Net-based model and trained to reconstruct a healthy rectum and mesorectum from prostate T2-weighted images (T2WI). The rectal anomaly maps were generated from the difference between the original rectal and reconstructed pseudo-healthy slices. The derived anomaly maps were used in the downstream tumor segmentation tasks by fusing them as an additional input channel (AAnnUNet). Alternative methods for integrating rectal anatomical knowledge were evaluated as baselines, including Multi-Target nnUNet (MTnnUNet), which added rectum and mesorectum segmentation as auxiliary tasks, and Multi-Channel nnUNet (MCnnUNet), which utilized rectum and mesorectum masks as additional input channels. As part of this study, we benchmarked nine models for rectal tumor segmentation on a large multicenter (num = 705) dataset of preoperative T2WI and nnUNet outperformed the other eight models on the external test. The MTnnUNet demonstrated improvements in both fully-supervised and mixed-supervised settings where human-annoated tumor masks and AI-generated rectum and mesoretum masks were used compared to nnUNet, while the MCnnUNet showed benefits only in the setting where mixed-supervision were used. Importantly, anomaly maps were strongly associated with tumoral regions, and their integration within AAnnUNet led to the best tumor segmentation results across both settings. The effectiveness of AAnnUNet demonstrated the value of the anomaly maps, indicating a promising direction for improving rectal tumor segmentation and model robustness for multicenter data.