Have you seen my "'Engineering Discussion" question about cleaning up an existing AI system when fake or inadequate information has been used during it's  training session?

Do you read Medical research articles published in the Lancet, JAMA or NEMJ about problems with the new AI systems?

One of my "work related jobs" before I fully retired was in "Configuration management" which involved products, processes and software.

The major problem was trying to determine if a proposed change was "Major" or "Minor", from a traceability viewpoint. Software changes proved very difficult in this respect. 

I recently see a lot of medical equipment recalls (for example reset glitches) because of software updates.

Peter 

Thanks Peter

WE are doing a few profiling exercises on specific sectors under the IEEE AI Ethics programme but have not done for eHealth or Medical applications yet. I do not read medical research articles specifically.

I am a fan of Config Mgt, preach and apply it in my work. appreciate your comments and observations however. hope we can convince the IET to embark on some action in this context.

cheers

Alas it does not seem possible to edit the message online.

If you're the author of a post you should be able to edit it using the "More" option underneath it.

   - Andy.

thanks for the guidance on editing Andy. Just managed it!

cheers

To me, medical related software system problems are the most "mission critical"  (as NASA would say - I live on the Space Coast) application.

Recently a company withdraw from the LVAD business because of software problems.

By the way LVAD is currently the "go to" bridge technology if one can not get a heart transplant. If the system stops you die!

Peter Brooks

Palm Bay