Research on Improving the Matching Efficiency between Cancer Patients and Clinical Trials Based on Machine Learning Algorithms

Abstract

Clinical trials offer critical opportunities for cancer patients to access novel treatments. However, the current trial matching process is often time-consuming, labor-intensive, and limited by fragmented data and manual screening. This study explores the application of machine learning algorithms to optimize the matching of cancer patients to clinical trials. By constructing structured representations of both patient profiles and trial eligibility criteria, and applying a combination of classification and similarity models, the system efficiently estimates match probabilities. Supplemented by natural language processing, feature extraction, and physician feedback mechanisms, the approach integrates automated recommendations with expert validation in a "model-assisted, human-in-the-loop" workflow. Case analyses demonstrate that this framework achieves high accuracy and significantly improves matching speed, providing effective support for personalized oncology care.