A HYBRID CLASSIFICATION AND GAME-THEORETIC STRATEGY FOR OPTIMIZING PATIENT RECRUITMENT IN CLINICAL TRIALS

Abstract

The urgent need for better patient recruitment in clinical trials, addressing issues like delays and excessive costs, makes this study essential. We want to improve patient outcomes, progress healthcare research, and increase trial efficiency by using a categorisation model and a game theoretic method for clinical trial settings. In the past, manual or semi-automated techniques including ads, referrals, and eligibility screening using databases have been used to recruit patients for clinical trials, despite the fact that this process is essential to the advancement of medical research. Trial completion is often delayed by these methods' inefficiencies, which include exorbitant expenses, drawn-out schedules, and inadequate participant matching. These difficulties show how conventional approaches fall short of the increasing need for effective and efficient hiring practices. In order to overcome these drawbacks, this paper suggests a sophisticated AI-driven solution that uses a Random Forest Classifier in a hybrid architecture. Through a stacking classifier, the suggested model combines Random Forests, Decision Trees, Support Vector Machines, and Logistic Regression. To improve prediction accuracy, autoencoders are used for feature extraction. Furthermore, via strategic interaction, a game-theoretic strategy involving patients, clinical investigators, and research firms optimises recruiting techniques, offering equilibrium-seeking solutions for stakeholder payoffs that are balanced. By simplifying the hiring process, this AI-based strategy provides a game-changing answer to the inefficiencies of conventional hiring practices and greatly advances the area of healthcare research.