AI-based Comprehensive Medical Assessments

Meta-machine Learning and Explainable AI: Performance Prediction of Medical Students in Serial Comprehensive Medical Assessments

introduction:

Serial comprehensive medical assessments are widely used in medical universities and academic hospitals to evaluate the competency of medical sciences students. However, these assessments can be time-consuming, expensive for universities, and stressful for students. To address these challenges, this study proposes a meta-machine learning (meta-ML) framework incorporating eXplainable artificial intelligence (XAI) to predict students' performance in serial comprehensive medical assessments. The framework aims to provide students, educators, and policymakers insights to identify at-risk students, determine specific courses requiring more attention, and develop targeted interventions. This study suggests that meta-ML models and XAI techniques can be reliable alternatives to comprehensive medical assessments. The findings can be valuable in identifying at-risk students and implementing evidence-based interventions to enhance students' academic achievements.

Code:

https://colab.research.google.com/drive/1l_T0EwjQGH_npusUcCcjwSYdxiqGLHIV?usp=drive_link

Keywords:

Artificial intelligence, Comprehensive medical assessments, Educational data mining, Explainable AI, Machine learning, and Medical licensing exams.

Corresponding Authors:

Toktam Dehghani; Email: [email protected];