Medical devices require robust quality assurance practices for AI models, which is a top priority, according to Troy Tazbaz, director of the Center for Devices and Radiological Health’s (CDRH) Digital Health Center of Excellence. His comments come just days after the Food and Drug Administration (FDA) released best practices for transparency in machine learning-enabled medical devices. Tazbaz detailed the agency’s approach to AI development and quality assurance in a recent blog post, stressing the importance of ensuring that AI models are accurate, reliable, ethical, and equitable.
Tazbaz highlighted several strategies for achieving these goals, including continuous monitoring of AI models before, during, and after deployment. This approach helps identify data quality issues and other problems before they impact the model’s performance. He emphasized that ongoing surveillance and proactive management are crucial for maintaining AI model integrity.
The FDA has been actively clarifying its stance on AI and machine learning in medical devices through various guidance documents and standards. In 2021, the agency collaborated with Health Canada and the U.K.’s Medicines and Healthcare Products Regulatory Agency to establish guiding principles for good machine learning practices. Recently, these agencies also issued principles on transparency, which include providing users with detailed information on how AI models generate their results.
FDA Defines Regulations for AI-Powered Medical Devices and Advocates Best Practices
In 2022, the FDA further defined which clinical decision support tools must be regulated as medical devices, specifically those that predict risks such as sepsis or stroke. Additionally, the agency issued draft guidance on predetermined change control plans, allowing developers to make post-market changes to AI models within pre-agreed parameters. The FDA is also co-leading an international working group on AI/ML-enabled medical devices with the International Medical Device Regulators Forum. This collaboration underscores the global effort to standardize and regulate AI in healthcare.
Tazbaz pointed out that AI has the potential to revolutionize patient care, enhance medical professional satisfaction, advance medical device research, and enable personalized treatments. However, he noted that the appropriate integration of AI in healthcare is crucial to realizing these benefits while minimizing risks and challenges.
To ensure AI technologies used as medical devices are safe and effective, the FDA’s Digital Health Center of Excellence advocates for adopting standards and best practices throughout the AI development lifecycle. This includes ensuring that data collection and quality align with the AI model’s intended use and risk profile. Tazbaz also called for the healthcare community to agree on common methodologies for informing users, including patients, about how AI models are trained, deployed, and managed.
FDA to Release New Guidelines on AI Quality Assurance in Medical Devices
The FDA plans to publish additional documents to further the discussion on AI quality assurance in medical devices. These future publications will address standards and best practices, the establishment of quality assurance laboratories, transparency and accountability, and comprehensive risk management strategies. According to market research firm Fortune Business Insights, the global digital pathology market, which includes AI-enabled medical devices, was valued at just over $1 billion last year. It is projected to grow to $1.15 billion in 2024 and reach $3.86 billion by 2032. This growth underscores the increasing importance of effective AI integration in medical devices.
AI’s role in the biopharma industry is also expanding, with a survey revealing that over 70% of major pharmaceutical companies and contract research organizations (CROs) have adopted digital pathology to advance drug research and development. This adoption is helping to accelerate the introduction of new therapeutics and improve research efficiency.
In conclusion, the FDA’s proactive stance on AI in medical devices, as outlined by Troy Tazbaz, highlights the importance of quality assurance and the adoption of best practices. By ensuring that AI models are accurate, reliable, ethical, and equitable, the FDA aims to enhance patient care and safety. As AI technology continues to evolve, the agency’s efforts to establish comprehensive standards and guidelines will play a critical role in integrating AI into the healthcare ecosystem effectively. These initiatives will not only improve diagnostic accuracy and efficiency but also foster innovation and collaboration in the field of digital health.
Resource: Food and Drug Administration, June 17, 2024
This article has been prepared with the assistance of AI and reviewed by an editor. For more details, please refer to our Terms and Conditions. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author.
