Ai-Driven Risk-Based Monitoring in Oncology Clinical Trials

View at publisher Summarise with VIRAT
Authors: Hemanth Dandu, Sumanth Singh
Cited by: 0
References: 9

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Risk-based monitoring is a relatively recent methodology of managing the monitoring of clinical trial data by using intelligent, data-driven methods instead of traditional data checks. This study is looking at the role that artificial intelligence might serve in risk-based monitoring of cancer clinical trials, which should have an increased focus on monitoring due to the variability of patients and the complexities of protocols. Predictive algorithms, Random Forest and XGBoost, were used on$\mathbf{1 3, 7 4 8}$oncology trials compliant with FDA regulations. Class imbalance had to be accounted for by using SMOTE, the Synthetic Minority Oversampling Technique. The application of SMOTE yielded an accuracy of$\mathbf{9 5 \%}$in monitoring with Random Forest; therefore, the methodology demonstrated improvement over conventional methods by 11.13 %. The paper also addresses difficulties associated with the prediction of adverse events and protocol deviation and illustrates how AI methods can drive improved quality monitoring, reduced burden for staff and stakeholders, and enable timely proactive responses to incidents to protect patients while assuring regulatory compliance.

Fields of Study & Subjects

Statistical Methods in Clinical Trials Pharmacovigilance and Adverse Drug Reactions Artificial Intelligence in Healthcare and Education

Publication details

Journal / source
International Conference on Communication, Computing and Electronics Systems (ICCCES)
Pages
1260–1266
Publication date
21 January 2026
Language
EN