Presentation
MDD8 - Can Artificial Intelligence Assist Medical Device Human Factors?
DescriptionHuman factors engineering (HFE), when employed in the development and assessment of medical devices, encompasses a range of activities, tasks, and documentation that demand both time and expertise to be executed properly.
Whether it be user research, task analysis, use-related risk analysis, usability testing, or protocol development, these specialized tasks are performed by HFE professionals with the shared goal of developing devices that are safe and effective when operated by a human user.
It may therefore appear unconventional for HFE professionals to consider that Artificial Intelligence (AI) tools enabled by computational power and algorithms could be beneficial in the domain of human factors. However, the emergence of such tools may have created new possibilities for optimizing the performance of the HFE professionals who utilize them in several ways:
-User-Centered Design: During the initial phases of user-centered design, AI may be able to assist in generating ideas for user interface layouts, controls, and feedback mechanisms that are intuitive and user-friendly. AI users can describe the target user population and context of use, and the AI tool can provide design recommendations.
-Usability Testing Scenarios: AI may be able to help craft realistic and relevant testing scenarios that mimic real-world use of the device, which is an essential part of effective usability testing.
-Heuristic Evaluation: AI can potentially perform heuristic evaluations of medical device interfaces, given principles, guidelines, and criteria.
-Task Analysis, PCA Analysis, and Use-Related Risk Analysis: AI tools can help break down medical device tasks into a step-by-step analysis, which is useful for generating use-related risk analyses, identifying potential use errors, and optimizing the design to mitigate them.
-Documentation and Labeling: AI may be able to assist in drafting user manuals, labeling, and instructions for use while ensuring the documents are clear, concise, and easy to understand, which is essential for usability and safety.
-User Feedback Analysis: AI has the potential to assist in analyzing and categorizing user comments and complaints to identify common issues or concerns, which can help prioritize design improvements.
-Regulatory Compliance: AI may be able to assist in providing information on relevant regulatory standards and guidelines related to human factors in medical devices, helping to ensure compliance with necessary regulations.
-Training and Education: AI can conceivably generate training materials or learning modules to educate end users on how to use a medical device correctly and safely.
This presentation will discuss the potential capabilities of AI tools such as ChatGPT in the context of human factors engineering. To explore the efficacy of these tools, we will demonstrate how they can be used to create a draft use-related risk analysis (URRA). To make this exercise more engaging, we will use ChatGPT to create the URRA draft live* in front of the audience.
*Note: We will prepare slides showing the same process in the case of technical difficulties.
We do not suggest that AI tools can, will, or even should replace human factors practitioners in the medical device development process. However, if there is value in these tools to support HF processes, they could be used as a complement to human expertise, not a replacement.
We will also discuss the limitations of AI tools as it applies to human factors activities. These will include considerations for analyzing images, PDFs, and highly unique/complex devices effectively as well as privacy/security concerns. Other concerns include complacency and over-reliance on AI tools.
Whether it be user research, task analysis, use-related risk analysis, usability testing, or protocol development, these specialized tasks are performed by HFE professionals with the shared goal of developing devices that are safe and effective when operated by a human user.
It may therefore appear unconventional for HFE professionals to consider that Artificial Intelligence (AI) tools enabled by computational power and algorithms could be beneficial in the domain of human factors. However, the emergence of such tools may have created new possibilities for optimizing the performance of the HFE professionals who utilize them in several ways:
-User-Centered Design: During the initial phases of user-centered design, AI may be able to assist in generating ideas for user interface layouts, controls, and feedback mechanisms that are intuitive and user-friendly. AI users can describe the target user population and context of use, and the AI tool can provide design recommendations.
-Usability Testing Scenarios: AI may be able to help craft realistic and relevant testing scenarios that mimic real-world use of the device, which is an essential part of effective usability testing.
-Heuristic Evaluation: AI can potentially perform heuristic evaluations of medical device interfaces, given principles, guidelines, and criteria.
-Task Analysis, PCA Analysis, and Use-Related Risk Analysis: AI tools can help break down medical device tasks into a step-by-step analysis, which is useful for generating use-related risk analyses, identifying potential use errors, and optimizing the design to mitigate them.
-Documentation and Labeling: AI may be able to assist in drafting user manuals, labeling, and instructions for use while ensuring the documents are clear, concise, and easy to understand, which is essential for usability and safety.
-User Feedback Analysis: AI has the potential to assist in analyzing and categorizing user comments and complaints to identify common issues or concerns, which can help prioritize design improvements.
-Regulatory Compliance: AI may be able to assist in providing information on relevant regulatory standards and guidelines related to human factors in medical devices, helping to ensure compliance with necessary regulations.
-Training and Education: AI can conceivably generate training materials or learning modules to educate end users on how to use a medical device correctly and safely.
This presentation will discuss the potential capabilities of AI tools such as ChatGPT in the context of human factors engineering. To explore the efficacy of these tools, we will demonstrate how they can be used to create a draft use-related risk analysis (URRA). To make this exercise more engaging, we will use ChatGPT to create the URRA draft live* in front of the audience.
*Note: We will prepare slides showing the same process in the case of technical difficulties.
We do not suggest that AI tools can, will, or even should replace human factors practitioners in the medical device development process. However, if there is value in these tools to support HF processes, they could be used as a complement to human expertise, not a replacement.
We will also discuss the limitations of AI tools as it applies to human factors activities. These will include considerations for analyzing images, PDFs, and highly unique/complex devices effectively as well as privacy/security concerns. Other concerns include complacency and over-reliance on AI tools.
Event Type
Poster Presentation
TimeTuesday, March 264:45pm - 6:15pm CDT
LocationSalon C
Digital Health
Simulation and Education
Hospital Environments
Medical and Drug Delivery Devices
Patient Safety Research and Initiatives