Body / AbdominalAI / InformaticsResearch

CECT-based 2PI system predicts HCC recurrence risk in dual-validation study

Cancer sciencetoday

A random survival forest model using a CECT-based 2PI imaging score predicted postoperative recurrence in solitary HCC (≤5 cm) with C-indexes of 0.76 (training), 0.69 (internal test), and 0.68 (external test). The system stratified risk using a threshold of ≥19.

  • The 2PI system combines seven CECT imaging features associated with pathological markers.
  • The RSF model was internally validated in a radiofrequency ablation cohort and externally validated at other centers.

Automated summary

RadPigeon summaries are original and for information only. They are not clinical advice.