Hepatocellular Carcinoma Segmentation on MRI: Five-Fold ATLAS Benchmarking, Phase-Wise External Stress Testing, and Multiphasic OpenSwissHCC Ablation
George Sgourakis · 2026
Read Hepatocellular Carcinoma Segmentation on MRI: Five-Fold ATLAS Benchmarking, Phase-Wise External Stress Testing, and Multiphasic OpenSwissHCC Ablation by George Sgourakis (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13611501.txt.
Description
George Sgourakis benchmarks an AI tool that outlines liver tumors on MRI scans in a 2026 paper. Automatic outlines help plan liver cancer care, but models often fail on new hospitals' scans. A careful public benchmark with outside stress tests was missing.
A standard self-configuring network trained on sixty public cases with five-fold checks. The frozen model then faced two outside collections across every scan phase. Accuracy on the training benchmark averaged a moderate overlap score. Performance shifted with scan phase and patient group, which the phase-wise tests expose. Scans without tumors checked the tool's false alarm rate separately. Adding a second contrast phase to training inputs was tested head to head. The study stresses auditability over headline scores.
Catalog record
| CREATOR | George Sgourakis |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 31 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |