AutoPET Challenge III: Testing the Robustness of Generalized Dice Focal Loss trained 3D Residual UNet for FDG and PSMA Lesion Segmentation from Whole-Body PET/CT Images

TL;DR

This study evaluates a 3D Residual UNet trained with Generalized Dice Focal Loss for lesion segmentation in PET/CT images, demonstrating promising results in the AutoPET Challenge 2024 with ensemble methods.

Contribution

The paper introduces the application of a 3D Residual UNet with Generalized Dice Focal Loss for robust lesion segmentation in PET/CT scans, validated through cross-validation and ensembling on the AutoPET dataset.

Findings

Achieved a mean DSC of 0.6687 in preliminary testing.

Ensembling improved segmentation robustness.

Provided open-source code for reproducibility.

Abstract

Automated segmentation of cancerous lesions in PET/CT scans is a crucial first step in quantitative image analysis. However, training deep learning models for segmentation with high accuracy is particularly challenging due to the variations in lesion size, shape, and radiotracer uptake. These lesions can appear in different parts of the body, often near healthy organs that also exhibit considerable uptake, making the task even more complex. As a result, creating an effective segmentation model for routine PET/CT image analysis is challenging. In this study, we utilized a 3D Residual UNet model and employed the Generalized Dice Focal Loss function to train the model on the AutoPET Challenge 2024 dataset. We conducted a 5-fold cross-validation and used an average ensembling technique using the models from the five folds. In the preliminary test phase for Task-1, the average ensemble achieved a mean Dice Similarity Coefficient (DSC) of 0.6687, mean false negative volume (FNV) of 10.9522 ml and mean false positive volume (FPV) 2.9684 ml. More details about the algorithm can be found on our GitHub repository: https://github.com/ahxmeds/autosegnet2024.git. The training code has been shared via the repository: https://github.com/ahxmeds/autopet2024.git.