FDA permits marketing of AI-based device to detect diabetes-related eye problems
1–10 of 104 posts
Re: FDA permits marketing of AI-based device to detect diabetes-related eye problems
#2Does the FDA conduct code reviews? And how do they guarantee that the code or the training data does not change over time without them knowing?
Re: FDA permits marketing of AI-based device to detect diabetes-related eye problems
#3Does the FDA conduct code reviews? And how do they guarantee that the code or the training data does not change over time without them knowing?
They generally treat compute devices as black boxes. Performance is all that matters. But once an artifact is submitted, it's locked down. That and only that will be approved. You change a resistor, an if-then, it goes back to FDA.
Re: FDA permits marketing of AI-based device to detect diabetes-related eye problems
#4I feel like this has the same promise as self driving cars - raise the floor for the quality of service while also experiencing random unexplained failures that its human supervisors fail to notice in time.
Re: FDA permits marketing of AI-based device to detect diabetes-related eye problems
#5The company is saying you don't need a specialist, but after bayes theorem (using 90%TN 87%TP and D(A)= 200,000 complication / 29,100,000 diabetes), the chance you have this condition after the machine says you do is 0.83%.
Re: FDA permits marketing of AI-based device to detect diabetes-related eye problems
#6Does the FDA conduct code reviews? And how do they guarantee that the code or the training data does not change over time without them knowing?
Seems like they are using this to get a proper referral to a specialist rather than using this as sole diagnosis. The code itself is probably 15 lines using Tensorflow or other framework I'm guessing, but could be wrong.
Re: FDA permits marketing of AI-based device to detect diabetes-related eye problems
#7I wonder what factors made this decision possible? I love the idea of automated diagnosis but the performance rates are 87% true positive and 90% true negative in the article. Seems a bit low.
Maybe people aren't getting diagnosed at very high rates? That would be a reasonable justification for deployment with somewhat less than perfect accuracy. Anyone have any insight?
Re: FDA permits marketing of AI-based device to detect diabetes-related eye problems
#8The company is saying you don't need a specialist, but after bayes theorem (using 90%TN 87%TP and D(A)= 200,000 complication / 29,100,000 diabetes), the chance you have this condition after the machine says you do is 0.83%.
Why are specialists immune to those same statistics?
Re: FDA permits marketing of AI-based device to detect diabetes-related eye problems
#9The company is saying you don't need a specialist, but after bayes theorem (using 90%TN 87%TP and D(A)= 200,000 complication / 29,100,000 diabetes), the chance you have this condition after the machine says you do is 0.83%.
I think you might have inverted the parameters. I think the numbers given are TN and TP.
Re: FDA permits marketing of AI-based device to detect diabetes-related eye problems
#10Hmm...I wonder how many patients actually diagnose themselves using the Internet (or their relatives do so)?
Source : personal experience.