Joint Liver and Hepatic Lesion Segmentation in MRI using a Hybrid CNN with Transformer Layers

TL;DR

This paper introduces SWTR-Unet, a hybrid CNN-transformer network that significantly improves liver and hepatic lesion segmentation accuracy in MRI and CT images, matching expert performance and aiding clinical workflows.

Contribution

The work presents a novel hybrid architecture combining CNN and transformer elements for improved hepatic lesion segmentation in medical imaging.

Findings

Achieved Dice scores of 98% for liver and 81% for lesions in MRI.

Demonstrated competitive results on CT data with Dice scores of 97% for liver and 79% for lesions.

Validated the method's accuracy against expert segmentations and across multiple imaging modalities.

Abstract

Deep learning-based segmentation of the liver and hepatic lesions therein steadily gains relevance in clinical practice due to the increasing incidence of liver cancer each year. Whereas various network variants with overall promising results in the field of medical image segmentation have been successfully developed over the last years, almost all of them struggle with the challenge of accurately segmenting hepatic lesions in magnetic resonance imaging (MRI). This led to the idea of combining elements of convolutional and transformer-based architectures to overcome the existing limitations. This work presents a hybrid network called SWTR-Unet, consisting of a pretrained ResNet, transformer blocks as well as a common Unet-style decoder path. This network was primarily applied to single-modality non-contrast-enhanced liver MRI and additionally to the publicly available computed tomography (CT) data of the liver tumor segmentation (LiTS) challenge to verify the applicability on other modalities. For a broader evaluation, multiple state-of-the-art networks were implemented and applied, ensuring a direct comparability. Furthermore, correlation analysis and an ablation study were carried out, to investigate various influencing factors on the segmentation accuracy of the presented method. With Dice scores of averaged 98+-2% for liver and 81+-28% lesion segmentation on the MRI dataset and 97+-2% and 79+-25%, respectively on the CT dataset, the proposed SWTR-Unet proved to be a precise approach for liver and hepatic lesion segmentation with state-of-the-art results for MRI and competing accuracy in CT imaging. The achieved segmentation accuracy was found to be on par with manually performed expert segmentations as indicated by inter-observer variabilities for liver lesion segmentation. In conclusion, the presented method could save valuable time and resources in clinical practice.