AI Systems in healthcare

overview

AI is gaining traction in the changing realm of healthcare—from diagnosis to treatments. With the growth of AI, which are systems capable of simulating human intelligence and thought processes, comes the need for standards that support safety, effectiveness, and trustworthiness of healthcare-based AI. As healthcare has unique considerations including sensitive health data and patient safety, standards that support informed regulatory measures are also essential.

ANSI also explored the role of public-private partnerships to enable standards development for AI in healthcare during their July 2024 brainstorming session.

Modern blueish image of a scientists using healthcare technology.

ENTERPRISE RISK MANAGEMENT AND AI COMMUNITY MEET-UP - THURSDAY, APRIL 28, 2022

Stakeholders indicated that networking, informal discussions, and making connections with diverse communities outside their own were extremely valuable, and continued dialogue was requested on the very important topic of AI in healthcare. ANSI hosted a community meeting up session which covered how organizations manage the risk of AI and how it fits into the overall enterprise risk management framework (including different aspects of AI technology risks, AI operational risks, AI vendor risks, managing AI controls assessments, governance of AI). Participants networked to discuss ten key questions:

What is enterprise risk management in your space?
How do we manage risk of AI as part of overall enterprise risk management (have enterprise risk management, but AI is new part) Look at different aspects of AI technology, the operational impact of AI (risk analysis, separation of duties, etc.)? What is the same, what is different?
Are traditional risk management approaches appropriate for AI? What risk management approaches are suitable for managing the risk of AI use?
What are the challenges or changes to governance of systems with the introduction of AI?
How do you address the relationship of the IT operating environment and the use of AI tools, including user interactions with the systems and the physical and logical environment in which they are used?
What are the privacy risks with the use of AI applications and are there strategies to manage these risks? How can privacy risk be managed in a way that supports the validity of the data sets of AI-enabled systems while protecting individual privacy?
What is the relationship of data privacy to trustworthiness of AI-enabled systems?
Where do you see issues of overtrust and/or undertrust in AI in healthcare and are there techniques to manage them?
How do you measure and manage bias in AI?
Any recent examples of AI in healthcare going right or wrong that were surprising?

DATA STANDARDIZATION AS BUILDING BLOCKS FOR AI IN HEALTHCARE WORKSHOP - NOVEMBER 30 - DECEMBER 1, 2021

The workshop included interactive panels to explore:

Standards activities: Attendees reviewed trends, themes, insights, and patterns of standards and roadmaps that are being used or being developed, based on ANSI’s AI standards landscape scan and stakeholder survey, as well as standards organization plans.
Data quality, data measurement, and data management needs in healthcare AI: This included all phases of AI data, data quality, and data measurement in government, patient, provider, and industry use cases, and in future scenarios.
Gaps, overlaps, and opportunities for coordination: Participants discussed and identified opportunities arising from discrepancies between the needs and wants expressed in stakeholder scenarios and use case requirements, versus the existing and planned standards landscape.

The event built upon the Institute's earlier workshop to support progress in AI standardization. ANSI’s 2021 report, "Standardization Empowering AI-Enabled Systems in Healthcare" reflects feedback from a 2020 ANSI leadership survey and national workshop, and identifies opportunities for standardization to support data, trust, transparency, governance frameworks, and risk management in this rapidly-growing area.

Event details Agenda

AI SYSTEMS IN HEALTHCARE STAKEHOLDER WORKSHOP - SEPTEMBER 14, 2020

ANSI convened a virtual, open forum stakeholder workshop, Standardization Enabling AI Systems in Healthcare, to explore opportunities for progress through collaboration and standardization, to identify challenges, barriers, and gaps, and to discuss steps to optimize regulatory frameworks in relation to artificial intelligence (AI) systems in healthcare.

The results of a July 2020 ANSI survey on Standardization Empowering AI-Enabled Systems in Healthcare inspired the workshop discussions, which covered data, transparency and explainability, governance, and risk management. The goal was to assess the need for coordination of standardization and governance to meet expectations of safety, quality, responsibility, and risk to support AI-enabled systems in healthcare.

Access meeting report

Staff Contacts

Have a question or need help?

Michelle Deane

Senior Director, Standards Facilitation

Phone:
212.642.4884

Email:
[email protected]