P104: A Systematic Review on the Use of Artificial Intelligence in Pharmaceutical and Medical Device Regulatory Affairs
Poster Presenter
Cezar Jr. Manansala
Student
Arellano University Philippines
Objectives
The objective of the study is to give insight to what is presently happening in the world of AI in Pharmaceuticals and Medical Devices around the world. Given the complexity and extensive datasets inherent to global Regulatory Affairs, a systematic review was conducted to map the current state of AI
Method
We searched PubMed, Scopus, Research Gate and Google Scholar for studies related to the use of Artificial Intelligence in Regulatory Affairs using the keywords “Artificial Intelligence”, “Regulatory Affairs”, “Pharmaceuticals”, “Medical Devices” for Articles between January 2023 to October 2025
Results
For all the studies included for the determination of the current state of AI in Regulatory Affairs, 22 records were identified through PubMed, Scopus, Research Gate and Google Scholar whose publication date is from January 2023 to October 2025 using the search field, “Title” or “Abstract” and search items were “Artificial intelligence and Regulatory Affairs or “Pharmaceuticals” and “Medical Device”. After duplicates were removed, 18 were included and full text articles for review is 16. We divided all of the studies into the three main stages of Artificial Intelligence as to what AI Overview of Google has suggested either Artificial Narrow Intelligence, Artificial General Intelligence and Artificial Super Intelligence. Artificial Narrow Intelligence is where systems are designed for and excel at a single, specific tasks such as playing chess, recognizing faces or making recommendations. Artificial General Intelligence is where AI would possess human-like cognitive abilities, allowing it to understand, learn and apply knowledge across a wide range of tasks, similar to human being and Artificial Super Intelligence is where AI would surpass human intelligence in virtually every field, including scientific creativity, general wisdom and social skills. All of the studies we have found are still in Stage 1 Machine Learning and in terms of division of the studies either Pharmaceuticals or Medical Devices, majority of the study are into Pharmaceuticals (56%), Medical Devices (25%) and Combination of Pharmaceuticals and Medical Devices (13%).
Conclusion
The role of Regulatory Affairs in Artificial Intelligence encompasses registered pharmaceutical products marketing authorization and related changes of the following: Regulatory Intelligence, Regulatory CMC, Managing CMC Data for Regulatory Submission, Global Regulatory Approvals in Pharmaceutical Research and Development, Drug Submissions and Variations. In terms of Regulatory Approval process, the application of AI has vastly been involved in labeling, regulatory publishing, regulatory submission, data analysis and pattern recognition, natural language processing, predictive modelling, post marketing surveillance and at a speed, the processing of immense amounts of data the hallmarks of AI and Machine Learning is incomparable in terms of speed and accuracy. In medical devices, artificial intelligence enabled machine learning has seen a significant increase from 2014 to 2023. In the order of decreasing use, here are the sample applications of artificial intelligence in medical device fields such as Radiology, Cardiovascular, Neurology, Hematology, Gastro-enterology, Anesthesiology, Ophthalmic, Clinical chemistry, General and Plastic surgery, Microbiology, Pathology, Orthopedic, General Hospital, Dental. In the Philippines, application time takes approximately three (3) years before the Certificate will be released based on actual circumstances experienced by the industry as. As I have adopted a new structure to which to be added is an Optical Character Recognition (OCR), Natural Language Processing (NLP) and Automated Processing among others, the review process will only take 4 months based on my estimation. The human verification will be the last touch in the review process which is still essential. The review process can, therefore, hence be shortened in time and greater accuracy and precision in reviewing.
