Page 12 - ITPS-8-3

INNOSC Theranostics and
Pharmacological Sciences AI in medical device safety

Drug Administration (FDA). These devices help prevent including health care. Numerous studies highlight AI’s
cardiovascular complications and provide early warnings groundbreaking potential in promoting data-driven
to users. However, the advent of these technologies raises decision-making. AI’s ability to swiftly and accurately
34
concerns about security, ethical bias, accountability, and process vast amounts of data has made real-time, data-
clinical effectiveness in practical settings. These concerns driven decision-making possible. Advanced language
emphasize the need for regulatory frameworks that manage models, such as ChatGPT, are already being used in
these risks while ensuring fairness and transparency. 35 government sectors to improve operations, policy-making,
and public services such as emergency response and public
AI-driven post-market surveillance provides 41
innovative approaches to identifying adverse events health management. In materiovigilance, where timely
decisions are critical for patient safety, this application can
and device malfunctions at an early stage. For example, be expanded. AI-driven decision-making has been shown
a framework for monitoring AI tools used in breast to improve organizational performance, with big data-
cancer screening across clinical centers highlights the powered AI playing a significant part in the development
importance of surveillance in detecting potential software of AI capabilities within organizations. 42
malfunctions. As AI continues to develop, establishing
36
robust quality management systems and stringent post- Despite its considerable potential to improve decision-
market monitoring procedures will be crucial to ensuring making, AI presents several challenges that need to be
the safety and efficacy of medical devices throughout their addressed. Issues such as data privacy, ethical implications,
lifecycle. 37 and the risk of negative outcomes from poorly implemented
AI remain prevalent. 43,44 These concerns are particularly
3.4. AI-driven automation in regulatory reporting significant in the context of materiovigilance, given the
AI-driven automation in regulatory reporting holds sensitive nature of medical records and the possible impact
significant potential for reducing human error and on patient safety.
enhancing operational efficiency, particularly in the In conclusion, real-time, data-driven insights from
areas of materiovigilance and medical device safety. By AI have the potential to significantly improve decision-
streamlining data collection, analysis, and submission making in materiovigilance. However, for successful
procedures, organizations can improve compliance and implementation, it is essential to carefully consider ethical
effectiveness in the medical device industry through the concerns, data privacy, and the development of robust
integration of AI technologies. 38 governance frameworks. 43
AI-driven automation has already shown considerable
promise in improving case reporting, data quality, and drug 4. Improving patient outcomes: Case
safety signal detection in the context of pharmacovigilance. studies of AI in materiovigilance
39
These advancements can also benefit in monitoring of Technologies such as AI and ML have shown significant
medical device safety. AI systems significantly reduce the promise in improving patient outcomes and advancing
need for manual labor by processing thousands of adverse materiovigilance processes. The utility of AI in enhancing
event reports each month, analyzing data, and interpreting patient safety and monitoring medical devices is
results at impressive speeds. 39 demonstrated by numerous case studies.
Despite the numerous benefits of AI, integrating these AI has the potential to improve intraoperative patient
technologies into regulatory reporting systems presents care in anesthesiology by continuously monitoring vital
several unique challenges. Two major obstacles are the lack signs and predicting complications, as exemplified by
of unified regulatory guidance and the availability of suitable the Hypotension Prediction Index algorithm. This AI
45
training data for ML models. In addition, while the goal of application enables anesthesiologists to optimize medication
39
complete automation is appealing, it should be approached dosages, reducing side effects and increasing effectiveness.
with caution, as stated in pharmacovigilance. A collaborative Similarly, smartwatches with AI capabilities have proven
approach that combines technical expertise with intelligent effective in detecting cardiac arrhythmias by continuously
technology must be prioritized, aiming to augment human tracking heart activity. One case study reported the use of a
capabilities rather than completely replace them. 40 smartwatch to identify atrial fibrillation in a young patient,
highlighting the potential of wearable AI technology for
3.5. Enhancing decision-making with AI-powered early diagnosis and intervention. Furthermore, during
46
insight the COVID-19 pandemic, CV and AI-driven predictive
AI-powered insights are increasingly being employed analytics were used to facilitate remote care, diagnosis, and
to improve decision-making across various fields, screening. These applications helped minimize physical
22

Volume 8 Issue 3 (2025) 6 doi: 10.36922/itps.6204