Study of a Computer-Aided Tool to Improve Predictions for Patients Undergoing Simultaneous Surgery for Colorectal Liver Metastases

What this Trial Is About

Researchers are evaluating a new computer-aided prognostic prediction tool designed to help doctors assess risks for patients with colorectal liver metastases CRLM. Colorectal cancer is a major cause of cancer-related deaths worldwide, and 20-30% of patients have liver metastases at diagnosis, which often leads to poor outcomes and high recurrence after surgery. This study aims to see if the tool improves doctors ability to predict postoperative recurrence and mortality for patients undergoing simultaneous resection of primary tumors and liver metastases. The study involves 12 physicians who will review 166 retrospective patient cases with CRLM. These doctors are split into two groups and will assess each case twice once without the tool and once with the tools assistance, separated by a washout period. The tool uses Random Forest models combining demographic, clinical, laboratory, and genetic data to predict 1-, 3-, and 5-year risks of recurrence and death. The main goal is to measure improvement in prediction accuracy for 3-year postoperative mortality, with additional outcomes including accuracy at other times, sensitivity, specificity, consistency between doctors, confidence in decisions, and evaluation time. Participants in this study are the physicians who will use the tool to evaluate the cases. Researchers will analyze their predictions and compare performance with and without the tool. The study will measure outcomes using the area under the receiver operating characteristic curve AUC-ROC and other statistical tools over time. This observational study is conducted by the Cancer Institute and Hospital, Chinese Academy of Medical Sciences, and aims to support better personalized decision-making for CRLM patients undergoing surgery.

Multi-Reader Multi-Case Trial Evaluating Computer-Aided Tool for Prognostic Prediction of Colorectal Liver Metastases