Artificial intelligence: 4 questions you should have an answer to

The use of artificial intelligence (AI) helps to simplify complex work steps in many medical areas. As an example, it assists in the diagnosis of disease patterns or has a supportive effect on its treatment.

However, as in all areas of medical technology, manufacturers must also fulfil specific regulatory requirements in order to enable the use of AI for medical products.

Four questions you should definitely think about before starting development:

Quantity
1. do I have access to enough high quality data?
2. Quality
2. does my data also cover all cases to be considered for the AI?
Requirements
3. can I fulfil the regulatory requirements and standards?
Traceability
4. is the result of the AI-based medical device still comprehensible by users later on?

What assumptions underlie the questions?

Artificial intelligence is only as intelligent as the data on which it is based. If an AI receives faulty or subjective data, the system is no longer reliable either. It is therefore particularly important to check the data sets it contains.

Not only the quantity of data is crucial. For an AI to be able to act meaningfully, it needs data for all decision cases that can potentially have an influence.

The ISO/IED TR 29119-11 and TR 24027 to 24029 standards currently apply to AI-based systems in Europe. An Artifact Intelligence Act (AIA) for the European Union is also currently in preparation.

In addition, there is a questionnaire from the Interest Group for Notified Bodies (IG-NB) on artificial intelligence in medical devices.

As far as an AI qualifies as part of the software of a medical device or is part of a medical device itself, the reliability, safety and usefulness of the AI must be classified and proven according to the MPDG.

Worth reading

What do manufacturers have to consider when approving AI-based medical software? And how do the regulatory requirements of the European market differ from those in the USA?

Read more about the topic in the blog “AI in medical devices – a regulatory perspective“.

Are you planning a project with artificial intelligence? Or would you like to find out about your options for developing a medical device with AI?

Feel free to contact us. We support you in regulatory affairs, quality assurance measures through to post-market surveillance.

Artificial intelligence: 4 questions you should have an answer to

Sounds interesting? Share our post