Guiding Principles for Conducting Clinical Trial for Machine Learning-enabled Medical Devices (MLMD)

The Ministry of Food and Drug Safety (MFDS) in Korea and the Health Sciences Authority (HSA) in Singapore have jointly released guiding principles for conducting clinical trials for machine learning-enabled medical devices (MLMD). These principles are designed to ensure the safety, effectiveness, and ethical conduct of clinical studies involving such devices.

Key Areas Covered in the Guidelines

• Clinical Trial Design:
  • Essential for ensuring the validity, reliability, and ethical conduct of the trial, addressing challenges specific to MLMD studies.
• Patient Selection:
  • Critical for the validity and generalizability of trial results, ensuring the selected cohort accurately represents the intended patient population.
• Test Dataset Selection:
  • Necessary for producing relevant and applicable data, ensuring that the trial outcomes reflect real-world applicability and robustness.
• Clinical Reference Standards:
  • Objective, pre-determined benchmarks used for comparison and assessment, often based on established clinical guidelines, providing a clear frame of reference.
• Primary Endpoints:
  • The main outcome measured in the trial, which directly determines the effectiveness and safety of the ML-enabled medical device.

Overall Aim

These guiding principles address the unique challenges faced in conducting clinical studies for MLMDs, ensuring rigorous evaluation while safeguarding the well-being of patients. They aim to promote effective clinical trial designs that generate reliable data, fostering the safe and ethical introduction of machine learning-enabled devices into clinical practice.

Region: Singapore, South Korea
Region: https://www.mfds.go.kr/eng/brd/m_40/down.do?brd_id=eng0011&seq=72637&data_tp=A&file_seq=1