Chinese researchers reported a machine-learning decision system to predict distant metastasis in a rare, aggressive primary liver cancer form using retrospective data. Published in Cancer Cell International, the study describes a model built and validated by investigators led by Lin Xu, Rongqiang Liu, and colleagues. The work focuses on metastasis risk assessment—an operational clinical question where earlier triage can influence staging, treatment selection, and trial enrollment. By translating imaging patterns into prognostic outputs, the approach aims to reduce the time between diagnostic workup and clinical decision-making. For oncology-focused biotech, the immediate relevance is how radiology-derived algorithms could become companion tools for pipeline assets, especially where trial eligibility depends on risk categories that are difficult to estimate quickly.