Day Four – 13 June 2024Thursday, 13 June 2024 - CEST (Cent Europe Summer, GMT+2)

• Identify and discuss appropriate testing methods for substances like creams, gels, fillers, etc.
• Handling chemical analysis for water-absorbent polymers

• The connections between biocompatibility and cleanliness of devices

• Discuss how standardizations differ from organization to organization

• Implementing of industry wide standardization practices

• Particulate testing within cleanliness validation and verification

• Current regulations relating to SaMDs and IVDs
  • Where is it covered in the MDR & IVDR?
  • Link to IEC standards and other legislation
• On-going work
  • Which standards are in the pipeline for update or introduction?

• Defining Artificial Intelligence and Machine Learning – what classifies?
• Overview of legislation relating to AI/ML
  • Where does the MDR, IVDR, other regulations, and standards link in?
• Future play – which pieces of legislation are on the horizon?

As the EU Artificial Intelligence Act (AIA) is expected to enter into force in 2024 and applicable in 2026, planning for its implementation should start now. The horizontal nature of the AIA means that it will interplay with the Medical Devices Regulation (MDR), where tensions and inconsistencies are likely to emerge. In particular, the AIA requires a single conformity assessment procedure for AI-enabled medical devices covered under the MDR. Therefore, for developers of AI-enabled medical devices, the AI Act poses several challenges, including additional obligations related to testing, data governance, post market monitoring and documentation. During this session, we will identify overlaps regarding the combined AI-MD conformity assessment and discuss the significant impact on the medical devices industry as well as ongoing debates. This session aims to enhance participants' comprehension of the conformity assessment procedures pertaining to AI-enabled Medical Devices under the Medical Device Regulation (MDR), while also addressing cybersecurity considerations.

Learning objectives:

• Have a better understanding of how the AI Act interlinks with the existing medical device regulation.

• Have a better understanding of how the MDR conformity assessment process will be different, due to the application of the AI Act for AI-enabled Medical Devices.

• Have a better understanding of what actions to take when Medical Devices is AI enabled and certified prior to the application of the Artificial Intelligence Act.

• Consideration of cybersecurity requirements under the MDR, which entails a comprehensive assessment of potential vulnerabilities and safeguards to ensure the integrity and security of AI-enabled Medical Devices throughout their lifecycle

• Overview of competitive landscape

Overview of regulatory environment

• Regulatory Authorities
• Device Classification
• Product Registration requirement
• Labeling requirement
• GxP requirement
• Post-market requirement

• Qualification of in vitro testing methods for sensitization and irritation
• Discussing regulatory hurdles
• Uncovering non-animal testing methods***

• A comparison of varying testing conditions used in biological testing for irritation.
• Potential contributing factors to failures in testing outcomes.
• Contextualising unfavourable results in biological testing for clinical relevance.

• What standards are developed at ISO / IEC level regarding AI
• How are ISO / IEC standards transposed into the EU landscape
• What standards developed at CEN / CENELEC level and how do these address AI Act requirements
• What are the implications on the technical documentation and quality management systems of manufacturers

• Broadening scope
  • Analysing the latest SNAIG report including some national regulation
  • Recommendations for works of IEC TC 62, JTC1SC42, ISO TC210 and ISO TC 215
  • Architecture of new standards in the Area of AI.
• IEC 63450
  • Timelines and scope of the IEC 63450 ED1 Testing of Artificial Intelligence / Machine Learning-enabled Medical Devices
  • Showing the line difference between IEC 63450 and PT 63521 “Machine Learning-enabled Medical Device – Performance Evaluation Process”
• Review of the Standard recommended by SNAIG and their impact on a QMS System for SAMD
  • BSI / AAMI 34971:2023, Application of ISO 14971 to machine learning in artificial intelligence
  • ISO/IEC TS 4213:2022, Information technology – Artificial intelligence – Assessment of machine learning classification performance
  • IEC 60601-4-1 Medical electrical equipment – Part 4-1: Guidance and interpretation – Medical electrical equipment and medical electrical systems employing a degree of autonomy (Scope: medical electrical equipment (MEE) and medical electrical systems (MES))

• Current documentation and regulations for SaMD and AI/ML under the FDA
  • Upcoming amendments and publications
• Routes to market
  • Utilising local contacts
  • Help available
• Future of US SaMD landscape

• Overview of competitive landscape

Overview of regulatory environment

• Regulatory Authorities
• Device Classification
• Product Registration requirement
• Labeling requirement
• GxP requirement
• Post-market requirement

• Breathing gas pathway biocompatibility for manual resuscitator
• Planning for biocompatibility in development of gas pathway device
• Testing strategy based on device design and intended use
• Learnings from MDR

Navigate the challenges surrounding the combined aspects of extractables and leachables, as well as chemical characterizations for combination products.

Patients, healthcare providers, payers, and regulatory bodies are increasingly demanding that healthcare organizations provide robust clinical data to prove their products’ clinical safety, performance, and economic value. Therefore, generating, analyzing, and distributing data in an effective manner is critical to pharmaceutical and medical device manufacturers. This can be a costly endeavor, with clinical trials costing millions of dollars and taking years to complete. As a result, effective data management and real-world evidence is becoming increasingly critical to the healthcare industry as companies look to provide the critical evidence they need to support their products.

This session will explore the challenges of siloed evidence collection and what steps organizations can take towards optimizing literature evidence management, fostering cross-functional collaboration with internal and external teams. We will also dive into the benefits of centrally storing and managing evidence, minimizing redundant efforts, reducing the time required to complete literature reviews, and accelerating regulatory compliance.

Key Learning Objectives:

• Overcoming challenges related to the evidence gathering process
• Establishing a corporate-wide centralized evidence management strategy
• Determining appropriate levels of data access and security to allow effective collaboration while safeguarding sensitive information
• Implementing and encouraging data reuse across multiple teams and business units
• Centralizing evidence collection and sharing practices to streamline regulatory compliance

• Latest update on the implementation of the AI Act
  • How does the AI Act interact with both the MDR and IVDR?
• GDPR
  • Difference in obligations based on risk level
• Putting theory into practice
  • Application timelines
  • Impact on regulators, notified bodies and manufacturers
• Forward thinking
  • Can we expect further regulations as AI develops?

• Overview of competitive landscape

Overview of regulatory environment

• Regulatory Authorities
• Device Classification
• Product Registration requirement
• Labeling requirement
• GxP requirement
• Post-market requirement

• The role of simulated use extraction studies in chemical characterization testing

• Simulation of physiological tissues, with a focus on blood

• Review of recently published comparison data between blood, saline, and alcohol/water mixtures for medical device extraction

• Legal and regulatory consequences of not getting your biocompatibility right under the Medical Devices Regulation

• Biocompatibility and product liability (Product Liability Directive and proposed Product Liability Directive Amendment)

• National enforcement trends

• Biocompatibility and legacy devices

• What are the key factors to address in an AI/ML device submission?
• Transparency with respect to AI/ML performance and data used to train / validate the model.
• Using best available methods for reference dataset
• Ensuring generalizability to the intended patient population
• Post-market considerations with respect to monitoring AI/ML performance and regulatory oversight

• Navigating the interplay
  • Where does the MDR and EU AI Act crossover?
  • Do expectations for industry align in both regulations?
  • How are Notified Bodies and industry navigating the interplay and how are they preparing for compliance?
• Ensuring compliance: how will conformity assessments look with the introduction of the EU AI Act?
• Moving forward
  • Best practices for navigating multiple regulations

• The Healing Power of Words: Large Language Model’s Potential in Healthcare
• Why Large Language Models Often Fall Short of Medical Device Legislation
• The Keys to Realizing the Potential of Large Language Models in Healthcare
• Unveiling the Impact of the European AI Act's Influence on Large Language Models for Medical Devices

• Overview of competitive landscape

Overview of regulatory environment

• Regulatory Authorities
• Device Classification
• Product Registration requirement
• Labeling requirement
• GxP requirement
• Post-market requirement