Enhancing Trust in AI for Healthcare: New BSI Guidance Released

BSI Publishes Guidance to Enhance Trust in AI for Healthcare

In an effort to strengthen digital trust in artificial intelligence (AI) products used for medical diagnosis and treatment, the British Standards Institution (BSI) has introduced new high-level guidance. This guidance, entitled Validation Framework for the Use of AI within Healthcare – Specification (BS 30440), seeks to enhance confidence among healthcare professionals, clinicians, and providers regarding the safe, efficient, and ethical development of AI tools.

As the global discourse surrounding the ethical application of AI evolves, this auditable standard focuses on products designed specifically for healthcare interventions, diagnoses, and health management. Jeanne Greathouse, Global Healthcare Director at BSI, emphasized the relevance of this standard for organizations in the healthcare sector and beyond, noting the transformative potential of AI as it increasingly integrates into healthcare practices.

Greathouse stated, “With the emergence of innovative AI tools, clinicians can leverage algorithms that efficiently analyze vast amounts of data to make diagnostic decisions, preventing and treating diseases, ultimately enhancing patients’ quality of life.” The projections indicate that the global healthcare AI market may exceed $187.95 billion by 2030. However, healthcare providers may encounter hurdles in evaluating AI products due to budget constraints, time limitations, or a deficiency in in-house technical expertise.

The BS 30440 specification aims to support decision-making processes by offering criteria for assessing healthcare AI products. This includes considerations like clinical benefits, performance metrics, safe integration into clinical practices, ethical implications, and equitable social outcomes. The standard encompasses a broad array of healthcare AI products, from regulated medical devices and imaging software to AI-powered chatbots and home monitoring systems.

This specification was developed with input from a diverse group of experts, including clinicians, software engineers, AI specialists, ethicists, and leaders in healthcare. It distills existing literature and best practices into an auditable framework, facilitating conformity with AI systems. Healthcare organizations can require BS 30440 certification during procurement to ensure compliance with established standards.

Scott Steedman, Director General for Standards at BSI, remarked, “This new guidance can foster digital trust in innovative tools that promise substantial benefits for patients and the professionals enabling their care. As AI continues to evolve, it is crucial to have confidence in the tools being developed, especially within healthcare.”

The specification addresses the demand for a cohesive validation framework for AI development and clinical evaluation in healthcare, building on initial pilots at Guy’s and St. Thomas Cancer Centre, refined through stakeholder discussions. With this guidance, BSI aspires to instill confidence in AI technologies used in healthcare, empowering clinicians, healthcare leaders, and patients to make informed, ethical decisions to enhance patient care and societal welfare.

As the role of AI in healthcare expands, adherence to recognized standards will be paramount in ensuring the safe and effective integration of these technologies into medical practice.