What requirements apply to AI functions in medical software?

Introduction

The integration of Artificial Intelligence (AI) in medical software brings numerous benefits but also requires adherence to strict requirements. These requirements are crucial to ensure the safety and effectiveness of the software and to gain user trust.

Technical Requirements

1. Safety

The safety of AI functions is of utmost priority. This includes both data security and the security of the algorithms themselves. The software must be designed to pose no risks to patients or users. This also includes protection against cyberattacks and data loss.

2. Effectiveness

The effectiveness of AI algorithms must be demonstrated through clinical studies. This means that the software must be capable of delivering the desired outcomes and meeting clinical requirements. Continuous performance monitoring is necessary to ensure that the software remains effective even after market launch.

Regulatory Requirements

1. Compliance with Standards

Medical software that integrates AI functions must comply with applicable standards and guidelines. This includes the EU Regulation on Medical Devices (Regulation (EU) 2017/745), which sets specific requirements for the safety and effectiveness of medical devices.

2. AI Regulation

Additionally, the requirements of the AI Regulation (Regulation (EU) 2024/1689) must be observed. This regulation specifies the criteria for the development and use of AI systems in various application areas, including medicine.

Conclusion

The requirements for AI functions in medical software are complex and require careful planning and implementation. Compliance with technical and regulatory requirements is essential to ensure the safety and effectiveness of the software and to secure user trust. Continuous validation and monitoring of AI functions are crucial to guarantee the quality of medical software in the long term.