TriALS: Triphasic-Aided Liver Lesion Segmentation Benchmark in Non-Contrast CT
This benchmark addresses the lack of annotated NCCT data for liver lesion segmentation, which is critical for low-resource settings where contrast agents are unavailable, but the results highlight that current methods still fall short on NCCT.
The TriALS challenge introduces the first annotated benchmark for liver lesion segmentation on non-contrast CT, using a multi-centre dataset of 150 cases. The top method achieved 0.754 Dice on venous-phase CT but dropped to 0.57 on NCCT, with external validation showing up to 28% improvement over off-the-shelf models, yet a persistent performance gap remains.
Automated segmentation of liver lesions on non-contrast computed tomography (NCCT) is clinically important but fundamentally challenging, particularly in low-resource settings across Africa and Asia where contrast agents are frequently unavailable. Progress has been limited by the absence of annotated NCCT benchmarks. Here we describe the TriALS challenge for automated liver lesion segmentation under contrast-limited conditions, supported by a multi-centre dataset of 150 cases with four-phase CT acquisitions (600 volumes) from Egyptian and Chinese institutions. Algorithms were evaluated on 70 cases from three institutions, including an independent external cohort. The top-performing method achieved a mean venous-phase Dice of 0.754, consistent with human-level performance, yet dropped to 0.57 on NCCT. On external validation, the leading method outperformed off-the-shelf models by up to 28% in Dice on NCCT. Algorithm performance was most strongly predicted by training data scale and pre-training strategy. A cross-year comparison exposed a persistent perceptual barrier on NCCT that scaling pre-training alone cannot overcome. Data, annotations, and code are available at https://github.com/xmed-lab/TriALS.