ISUOG Women's Health Ultrasound Artificial Intelligence strategy
The International Society of Ultrasound in Obstetrics and Gynecology (ISUOG) recognises the significant opportunity of Artificial Intelligence (AI) technologies for improving Women’s Health. Still, this appeal must be balanced against the need for careful prior evaluation of their strengths and limitations and the appreciation of the necessary safeguards to protect patients and clinicians prior to implementation. Key to successful implementation of AI-based technology is establishing a responsible AI strategy. Thus, ISUOG is committed to the following:
- Support the responsible development and adoption of AI technology in women’s health: ISUOG is committed to supporting the ethical, transparent, accountable and evidence-based development and implementation of AI technologies in women’s health. Equity must be a guiding principle in the development and deployment of AI in women’s health, ensuring that tools are inclusive, unbiased, and responsive to the diverse needs of all individuals. We aim to foster a collaborative environment where AI innovation is guided by clinical expertise, patient safety, and ethical considerations, ensuring that new technologies enhance rather than compromise patient care. We will promote the awareness of the level of risk for each AI system depending on the scope of its impact on patient care: from low risk, performing administrative or non-critical tasks, to moderate risk where the AI assists clinicians in making decisions but still requires human oversight, and high risk when AI can make autonomous decisions that directly affect patient care and outcomes.
- Teach AI technologies in women’s health: ISUOG will lead in educating healthcare professionals about AI technologies in women’s health. We will create engaging training programs, workshops, and multimedia resources to increase understanding of AI applications, capabilities, and limitations, ensuring practitioners can make informed decisions about AI integration in clinical practice. We will train healthcare professionals to distinguish AI systems based on the level of risk they pose to patient safety, healthcare outcomes, and clinical decision-making.
- Define research priorities in AI in women’s health globally: ISUOG will lead efforts to define and advance global research priorities in AI for women’s health. By identifying key gaps, opportunities, and challenges, we aim to drive innovation and evidence generation, ensuring that AI research aligns with the most pressing clinical needs and benefits women worldwide. ISUOG is committed to upholding integrity, transparency, and efficiency to ensure responsible AI implementation.
- Develop teaching and learning in women’s health, through the adoption of appropriate AI technologies to maximise benefit to the patient:
ISUOG will integrate AI into its educational programs to enhance the teaching and learning experience in women’s health. By leveraging AI-driven tools and insights, we aim to empower healthcare professionals with advanced knowledge and skills, ensuring that AI adoption improves patient outcomes and facilitates more effective clinical decision-making. - Use AI to maintain and increase integrity and ethical standards in publishing: ISUOG will adopt AI technologies to enhance the integrity, transparency, and efficiency of its publishing processes. We will implement AI-driven tools to improve peer review, detect potential conflicts of interest, and uphold the highest standards of scientific accuracy and ethical conduct in all published work.
- Use AI to improve operational efficiency: ISUOG will embrace AI to streamline internal processes and enhance operational efficiency. By integrating AI into administrative functions, member engagement, and event management, we aim to optimise resource allocation and improve the overall experience for our members and stakeholders.
While AI has the potential to enhance women’s healthcare, it is important to approach its adoption with caution, responsibility, and an eye toward the broader implications. ISUOG will ensure that AI tools are used safely, ethically and responsibly in the service of better health for women. ISUOG aims to shape the future of AI in various aspects of ultrasound and other technologies in women’s health, from clinical practice to research and education.
Key ISUOG recommendations when considering the responsible implementation of a new AI technology:
- Transparency: The methods and datasets used to develop and validate the AI should be clearly described.
- Applicability: Ensure the AI training dataset is representative (demographics and ethnicity of the populations used to train and validate the AI tool).
- Risk assessment: Recognises the level of risk of the AI tool: low-risk (administrative or non-critical tasks, posing minimal harm if it fails) vs. moderate-risk (AI assists clinicians in making decisions but still requires human oversight) vs. high-risk (AI can make autonomous decisions that directly affect patient care and outcomes).
- Evidence and evaluation: Evidence of technical evaluation (diagnostic performance, impact on workflow, clinical safety, international standards, equity, user-specific, population-specific) of the AI tool.
- Impact assessment: Availability of an analysis of the likely impact in the setting where it will be implemented.
- Successful implementation needs to ensure it fits with local data governance and existing IT systems.
- Compliance with the guidelines and requirements set forth by relevant regulatory and professional bodies.
- Standard operating procedures need to be in place for a critical interpretation of AI findings, including discrepancies. They also need to ensure appropriate human oversight and continuous monitoring.
- Adequate training: A training package that outlines the potential risks should be available for users.
- Availability of support during implementation: The option of seeking support from existing centres of AI expertise should be considered.
- Commitment to ongoing post-implementation surveillance.
- Attribution: Disclose the use of AI.
The ISUOG Board of Trustees would like to thank Professors Asma Khalil and Fatima Crispi, for all their work developing this position statement, and for their vision to embrace AI positively for the good of women’s health.