Reading Race: AI Recognises Patient's Racial Identity In Medical Images · HackerPapers discussions

Reading Race: AI Recognises Patient's Racial Identity In Medical Images — mentioned by kthejoker2

Sat, 11 Dec 2021 15:30:27 GMT

A conceptually related finding that computer vsion models can effectively classify patient race from significantly noised x-ray images https://arxiv.org/abs/2107.10356 > However, our findings that AI can trivially predict self-reported race -- even from corrupted, cropped, and noised medical images -- in a setting where clinical experts cannot, creates an enormous risk for all model deployments in medical imaging: if an AI model secretly used its knowledge of self-reported race to misclassify all Black patients, radiologists would not be able to tell using the same data the model has access…

Paper and discussion history. A mention is not an endorsement.

Reading Race: AI Recognises Patient's Racial Identity In Medical Images — mentioned by lostlogin

Tue, 14 Sep 2021 19:24:59 GMT

> they found it looks at the name of the patient on the border of the image or something similar. The images were anonymised. If they AI cheated, it’s not obvious how. The paper is interesting. https://arxiv.org/abs/2107.10356

Paper and discussion history. A mention is not an endorsement.

Reading Race: AI Recognises Patient's Racial Identity In Medical Images — mentioned by OneEyedRobot

Tue, 14 Sep 2021 14:50:03 GMT

They always call for 'further investigation'. Looking at this: https://arxiv.org/pdf/2107.10356.pdf My general impression (no more than that) is a whole bunch of people crowding into a paper. The paper is mostly applying trivial image processing functions and seeing how some software they don't understand is responding. The main aim is pearl-clutching about 'bias' rather than any kind of understanding. God knows what they're going to do when any medical exam includes some kind of deep dive into the patient's genetics. No surprises. It's the nature of the era.

Paper and discussion history. A mention is not an endorsement.

Reading Race: AI Recognises Patient's Racial Identity In Medical Images — mentioned by hgial

Tue, 14 Sep 2021 14:28:07 GMT

It might be helpful for folks to look at the blog post written by one of the authors: https://lukeoakdenrayner.wordpress.com/2021/08/02/ai-has-the... or the paper itself https://arxiv.org/pdf/2107.10356.pdf I see a lot of "oh it's probably just picking up on x y z" when x, y, and z are things they explicitly checked for: 1) "It's probably just the names or other metadata" – they only gave it pixel data to train on. To control for things like metadata overlaid on the image (e.g., a name written on the image) they divided the images into 3x3 sections and trained classifiers on each section…

Paper and discussion history. A mention is not an endorsement.

Reading Race: AI Recognises Patient's Racial Identity In Medical Images — HN submission

Sun, 22 Aug 2021 13:33:18 GMT

Reading Race: AI Recognises Patient's Racial Identity in Medical Images [pdf]

1 point; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Reading Race: AI Recognises Patient's Racial Identity In Medical Images — HN submission

Sun, 15 Aug 2021 19:27:02 GMT

AI Recognises Patient's Racial Identity in Medical Images

1 point; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Reading Race: AI Recognises Patient's Racial Identity In Medical Images — mentioned by bArray

Mon, 09 Aug 2021 14:11:06 GMT

Disclaimer: Worth noting that this paper is apparently pre-print [1] and was submitted to the Lancet, who are not particularly reliable in recent times (flaws hydroxychloroquine study [2], stamped out early investigation in COVID lab-leak possibility [3]). Regarding the blog and paper: > AI can trivially learn to identify the self-reported racial identity of patients to an absurdly high degree of accuracy If it is 'trivial' for a model to learn race, then what they are detecting is likely pretty obvious. In their paper they use 'private and public datasets' [4]. There is probably exactly…

Paper and discussion history. A mention is not an endorsement.