The Medical Imaging & Technology Alliance (MITA), a trade association representing manufacturers of medical imaging equipment, radiopharmaceuticals, contrast media, and focused ultrasound therapeutic devices, participated in the National Institute of Standards and Technology’s (NIST) third workshop to develop a voluntary Artificial Intelligence (AI) Risk Management Framework to address challenges unique to AI systems and encourage and equip different AI stakeholders to manage AI risks.

At the workshop, AI experts from across industries and NIST staff discussed the community’s feedback on the Framework and next steps toward improved AI risk management. MITA pointed to the medical device manufacturing industry as a leading example of a sector that has a successful history evaluating, mitigating, and controlling for risks at the forefront of modern technology—including AI.

Although regulatory requirements provide an important oversight mechanism, industry standards developed and iterated by medical device manufacturers continuously raise the bar for safe and effective use. In fact, a standards-based approach is responsible for the machine learning (ML) and AI techniques used in computer-aided detection software (CADe), computer-aided diagnosis software (CADx), and even devices such as the artificial pancreas, MITA says

“Risk management in medical device manufacturing is centered squarely on patient safety,” says Patrick Hope, executive director of MITA. “A standards-based approach has helped to establish the healthcare sector as one of the most trusted sectors in the country. However, the approach hinges on the experts within the healthcare industry coming together and establishing best practices and minimum requirements based upon state-of-the-art designs and technologies. As NIST continues to address risks introduced by AI, we believe the medical device industry’s approach to patient safety standards can serve as a model for AI risk management.”