I already immediately left a veterinarian for using AI as part of its xray diagnosis process, which may even be somewhat acceptable since computer vision is relatively mature. Fuck if I’m lasting 5 mins with a human doctor that utters the letters “AI.”
I can’t speak for veterinary, but in dentistry AI can be problematic because:
A) it really over-diagnoses - it’s very very sensitive, meaning that it identifies things that aren’t necessarily clinically relevant
B) it does not compare to previous radiographs, so it cannot give reasonable clinical judgement on whether decay is active or arrested.
It could be a helpful tool to give you a laundry list of places to check. However, I’ve used demo software and did not find it added anything for me, although I have 15 years experience. You still need to use your clinical judgement.
I do worry about the younger clinicians being over-reliant on it, as they have it pushed on them by multi-practice dental corporations.
Yep, the “laziness” of reliance was a big part of my immediate reaction. I’ve recently found out that veterinary clinics across the country, and especially in Texas, were aggressively moved on by investment capital in the years between now and my last dogs. The pet economy, including medical care, has long been seen as recession proof. Medical care in particular is seen as emotionally driven and ripe for greater exploitation, so of course the vultures moved in. Having AI flag everything, allowing you to show a customer tons of things to worry about and milk them for more diagnostics, a.k.a. more money, is the logical conclusion.
I already immediately left a veterinarian for using AI as part of its xray diagnosis process, which may even be somewhat acceptable since computer vision is relatively mature. Fuck if I’m lasting 5 mins with a human doctor that utters the letters “AI.”
Ya an LLM is very different than a vision based machine learning /trained visual model on something specific like x-rays.
Now, if its just a LLM looking at an xray image, that’s another story, and it could’ve been that too.
Exactly. Imagery is my field so them just saying “AI X-ray analysis” instead of something more specific or scientific didn’t inspire confidence.
I can’t speak for veterinary, but in dentistry AI can be problematic because:
A) it really over-diagnoses - it’s very very sensitive, meaning that it identifies things that aren’t necessarily clinically relevant
B) it does not compare to previous radiographs, so it cannot give reasonable clinical judgement on whether decay is active or arrested.
It could be a helpful tool to give you a laundry list of places to check. However, I’ve used demo software and did not find it added anything for me, although I have 15 years experience. You still need to use your clinical judgement.
I do worry about the younger clinicians being over-reliant on it, as they have it pushed on them by multi-practice dental corporations.
New fear unlocked
Yep, the “laziness” of reliance was a big part of my immediate reaction. I’ve recently found out that veterinary clinics across the country, and especially in Texas, were aggressively moved on by investment capital in the years between now and my last dogs. The pet economy, including medical care, has long been seen as recession proof. Medical care in particular is seen as emotionally driven and ripe for greater exploitation, so of course the vultures moved in. Having AI flag everything, allowing you to show a customer tons of things to worry about and milk them for more diagnostics, a.k.a. more money, is the logical conclusion.