The European Medicines Agency (EMA) has unveiled a comprehensive Artificial Intelligence (AI) Work Plan to guide the integration of AI technologies into regulating medicines across Europe through 2028. This initiative underscores EMA's commitment to harnessing AI's potential while ensuring patient safety and maintaining the integrity of regulatory processes.
EMA's AI Work Plan focuses on four pivotal areas:
Guidance, Policy, and Product Support: Developing and disseminating guidelines on AI applications throughout the medicinal product lifecycle. This includes providing continuous support for products in development and formulating policies that address the evolving landscape of AI in healthcare.
AI Tools and Technology: Establishing frameworks for the adoption of AI tools to enhance efficiency, data analysis, and decision-making within regulatory processes. Emphasis is placed on ensuring these tools comply with data protection legislation and uphold ethical standards.
Collaboration and Training: Building capacity and expertise among regulatory network staff, stakeholders, and partners to stay abreast of AI advancements. This involves fostering collaborations and providing training programs to equip personnel with the necessary skills to navigate AI technologies effectively.
Experimentation: Encouraging structured experimentation with AI applications to accelerate learning and gain insights into their practical implications. This approach aims to balance innovation with risk management, ensuring that AI deployments are both effective and safe.
The Big Data Steering Group developed this work plan, which received endorsements from the Heads of Medicines Agencies (HMA) in November 2023 and EMA's Management Board in December 2023.
In conjunction with the AI Work Plan, EMA released a Reflection Paper discussing the application of AI and machine learning (ML) across all stages of a medicine's lifecycle. This document offers considerations for medicine developers and marketing authorization applicants on the safe and effective use of AI/ML, emphasizing compliance with EU legal requirements, data protection laws, and ethical principles
Demonstrating a practical application of AI in medicine development, the EMA issued a Qualification Opinion on an AI-based tool designed to assess histological features in liver biopsies for Non-Alcoholic Steatohepatitis (NASH) and Metabolic Dysfunction-Associated Steatohepatitis (MASH) clinical trials.
The AIM-NASH tool assists pathologists in scoring liver biopsies, aiming to improve accuracy and reproducibility in evaluating disease activity and fibrosis stage. Following the draft Qualification Opinion for the AIM-NASH tool, EMA received public comments highlighting various aspects of its application. Stakeholders raised concerns about transparency, validation methodologies, and the potential biases of AI-driven histological assessments. Many experts emphasized the importance of comprehensive clinical validation before widespread adoption, ensuring that the tool does not introduce variability in trial results.
The AI Work Plan, along with related initiatives, sets a structured pathway for the responsible adoption of AI, ensuring that its benefits are realized without compromising ethical standards or patient safety.
References
Comments