Quality Management 1 answer

How to refine and standardise data quality control requirements for medical device algorithms?

Anonymous · Published March 04, 2026 · 1 comment
Data quality control is important for medical device product quality, especially for data-based algorithms. What efforts are made to standardise data quality control requirements in areas such as data collection, organisation, labelling, and dataset construction? How can these requirements be refined effectively?

Join the discussion. Leave a comment. Guest comments are welcome — add your email to get reply notifications.

Anonymous 3 months ago
Standardising and refining data quality controls is increasingly relevant for SaMD and AI/ML-based devices. Are there reference standards or industry best practices?
Reply to this comment

Discussion

1 Answer

Accepted answer Dr. Oliver Eidel · Founder & CEO, OpenRegulatory ·
We also design our software prototypes in Figma and keep all previous versions and drafts in a separate Figma board. For documentation, we link to the Figma board instead of copy-pasting wireframes into every document. Whenever there is a change, we duplicate the board or copy the relevant screens into a history board, add the change date and notes for traceability, and keep only the current mockups in the main board to avoid confusion for developers. This way, we have an archive of all versions and can easily reference or export them for audits if needed.

Join the discussion. Leave a comment. Guest comments are welcome — add your email to get reply notifications.

No comments yet. Be the first to share your thoughts.

Want to add your answer to this question?
Write an answer under your name by logging in or signing up, or post anonymously.

Still have a question? Ask a question here publicly — for free.

Or talk to one of our consultants — first calls are free. Check out our services and prices.

Looking to automate your regulatory work? Check out our eQMS, Formwork. Built for lean, founder-led companies. There’s a free version too.