Deep Learning-Based Segmentation of Peritoneal Cancer Index Regions from CT Imaging

TL;DR

This study develops and evaluates deep learning models, nnU-Net and Swin UNETR, for automated segmentation of radiological PCI regions from CT scans, aiming to replace invasive assessment methods.

Contribution

It introduces a deep learning approach for automatic segmentation of rPCI regions on CT, providing a non-invasive alternative to traditional methods.

Findings

nnU-Net achieved a Dice score of 0.82, close to interobserver agreement.

Swin UNETR scored 0.76, indicating competitive performance.

Challenges remain in segmenting right flank and small-bowel regions.

Abstract

Peritoneal metastases are currently assessed using diagnostic laparoscopy to determine Sugarbaker's Peritoneal Cancer Index (sPCI), which works by dividing the abdomen into 13 regions and scoring each region based on tumor size. A recent consensus study defined 3D regions to facilitate a radiological PCI (rPCI), providing standardized anatomical regions for imaging-based assessment. Despite its clinical value, sPCI is invasive and lacks a standardized imaging counterpart. In this study, we propose a deep learning-based approach to automatically segment the rPCI regions on CT. We evaluate nnU-Net and Swin UNETR on 62 CT scans with rPCI regions manually annotated by three clinical researchers and validated by two expert radiologists. Performance was assessed using five-fold cross-validation with the Dice Similarity Coefficient (Dice), 95th percentile Hausdorff distance and Average Surface Distance. nnU-Net achieved an overall Dice of 0.82, approaching interobserver agreement (0.88) and outperforming Swin UNETR (0.76), with remaining challenges primarily in right flank and small-bowel regions. These results demonstrate feasibility of automated rPCI segmentation, laying the foundation for non-invasive, imaging-based assessment.