AI Assisted Scientific Review: A future Scope

Abstract

The incorporation of Artificial Intelligence (AI) in the scientific and regulatory review procedure poses a transformational change in the evaluation of the pharmaceutical, biotechnology, and medical device products in terms of safety, efficacy and quality. Based on manual review of documents and human based judgment, regulators are increasingly applying AI tools, including machine learning, natural language processing and predicative analytics. The present paper discusses the way international regulators, such as the U.S. FDA, UK NICE, Health Canada, and others, embrace AI-based technologies as a better means of data screening, risk identification, and decision-making. It presents some concrete projects such as the ELSA tool developed by the FDA highlighting significant decreases in the time take to review and more safety issues revealed. There is also the comparison, in the paper, of the manual and the AI-assisted review procedures, as well as the challenges posed by the quality of the data and the transparency of the algorithm as it is experienced today, and the prospective path of the AI in regulatory science, such as the real-time decision support, global harmonization, and the post-market surveillance. The application of AI will shift, as it moves towards becoming an essential instrument within regulatory work processes by facilitating rapid, consistent, and patient-friendly scientific review.