Transparency of artificial intelligence/machine learning-enabled medical devices
Aubrey Shick, Christina M. Webber, Nooshin Kiarashi, Jessica Weinberg, Aneesh Deoras, Nicholas Petrick, Anindita Saha, Matthew C. Diamond
Abstract
Artificial intelligence/machine learning (AI/ML)-enabled medical devices (AI/ML devices, hereafter) offer exciting new opportunities to continue the advancement of healthcare. These innovative solutions can enable earlier disease detection, new insights into human physiology, improved and personalized diagnostics and therapeutics, and offer a unique ability to learn, adapt, and improve device performance as technologies and clinical contexts evolve or change 1 . AI/ML devices can harness the ever-growing amount of data available to actively learn from individuals using them during their lives, expanding the insights gained by decision makers, including healthcare providers, patients, caregivers, regulators, and payors 2 .