Identification of Known Use-Related Problems in Artificial Intelligence Medical Imaging Devices

Abstract

As Artificial Intelligence (AI) advances, it is included into more and more medical device systems to carry out these systems’ intended functions. Given AI’s ability to learn from real-world use and its capability to continuously improve performance, manufacturers of medical devices are utilizing AI to innovate their products to better assist health care providers. However, like application of other types of medical devices, the application of AI in medical care might pose different types of potential risks to the patients, the users themselves, and to the use environment. For manufactures to successfully ensure the safety of AI medical devices, it is crucial to identify known problems by investigating use-related, user interface, and user interaction in-cidents that have occurred in comparable medical devices. The objective of this study is to identify potential use-related problems of DeepView® Wound Imaging System that assesses the healing potential of thermal burn wounds by analyzing multispectral images with an ML algorithm. We use a variety of sources of information on reports and recalls of medical devices that are associated with deaths, serious injuries and mal-functions. After examining relevant reports and recalls, 19 use-related problems were identified. An in-depth analysis was then conducted, considering each identified use-related problem and determining how it relates to the specific features and functionality of DeepView® Wound Imaging System as well as identifying patterns and commonalities among them. The information gained from this analysis can be beneficial in enhancing the safety of AI medical devices by providing a deeper understanding of the reasons for failure, to avoid similar issues in the future, and ultimately to improve patient safety and public health.

Get full access to this article

View all access options for this article.

References

Food and Drug Administration. (2016). Applying Human Fac- tors and Usability Engineering to Medical Devices. Redtrieved January 8, 2023, from https://www.fda.gov/regulatory-information/search-fda-guidance-documents/applying-human-factors-and-usability-engineering-medical-devices

Food and Drug Administration. (2019). Proposed regulatory framework for modifications to artificial intelligence/machine learning (AI/ML)-based software as a medical device (SaMD).

Food and Drug Administration. (2021). Artificial Intelligence and Machine Learning in Software as a Medical Device. Retrieved January 8, 2023, from https://www.fda.gov/medical-devices/software-medical-device-samd/artificial-intelligence-and-machine-learn-ing-software-medical-dev

International Electrotechnical Commission. (2015). IEC 62366-1: 2015, Medical devices—Part 1: Application of usability engineering to medical devices. Geneva, Switzerland: International Organization for Standardization.

Jiang

Zhi

Dong

… & Wang

(2017). Artificial intelligence in healthcare: past, present and future. Stroke and vascular neurology, 2(4).

Pape

S. A.

Skouras

C. A.

Byrne

P. O.

(2001). An audit of the use of laser Doppler imaging (LDI) in the assessment of burns of intermediate depth. Burns, 27(3), 233-239.

Thatcher

J. E.

Squiers

J. J.

Kanick

S. C.

King

D. R.

Wang

… & DiMaio

J. M.

(2016). Imaging techniques for clinical burn assessment with a focus on multispectral imaging. Advances in wound care, 5(8), 360-378.