Advanced Tumor Segmentation in PET/CT Imaging: A Training Strategy Study with nnU-Net for AutoPET III

TL;DR

This study enhances tumor segmentation in PET/CT images using nnU-Net with optimized training strategies, achieving high accuracy and robustness across diverse data in the AutoPET III challenge.

Contribution

It systematically investigates training strategies for nnU-Net to improve tumor segmentation performance in multi-center PET/CT data.

Findings

Achieved Dice score up to 0.80 on test data.

Significant impact of training strategies on model robustness.

Ranked third in the AutoPET III challenge.

Abstract

Tumor segmentation in whole-body PET/CT imaging is crucial for precise disease evaluation and treatment planning. However, it remains challenging due to variability in lesion size, contrast, and anatomical distribution. Relying on manual segmentation makes the process time-consuming and prone to intra- and inter-observer variability. This work presents a whole-body tumor segmentation method developed for the AutoPET III challenge, where the goal is to build models that generalize across tracers and multi-center data. We employ the nnU-Net framework with a ResNet-based encoder as our baseline and systematically investigate the impact of training strategies, including intensity normalization, batch dice optimization, and data augmentation using CraveMix. Our experiments show that these strategies significantly influence model performance, particularly in reducing false positives and improving robustness to lesion variability. The best-performing configuration achieves a Dice score of up to 0.80 on the preliminary test phase, and our method ranked third in the AutoPET III challenge. The code is publicly available here.