If a doctor used a stethoscope with an app to diagnose me, I would get a second oppinion
I have a stethoscope modification. It’s a little device that goes right into the tubing. It’s primary use is an amplifier to make it easier to hear quiet sounds.
It looks perfectly normal other than the bulge.
It uses Bluetooth to transfer recordings to my phone.
But the stethoscope works normally otherwise you probably wouldn’t notice unless you knew what you were looking for.
Of course my software is not that fancy it doesn’t diagnose.
I’m also not a doctor. So you don’t have to worry about me.
Conveniently, you would have already gotten the doctor and the AI’s, no need to look any further! /s
If you want a second opinion feel free to ask the roomba.
The fridge would also like a word.
Audio recordings in this would be useful, but the rest just kills the product.
If done well, training AI on this kind of data could be a good thing. It could make say your smartphone (for example) tell you that you have a problem, be used as a first diagnostic and so on. Invaluable in countries where doctors are scarce for example.
If I understand correctly the patients consent is never asked before their data is collected, I could easily see this data being sold to advertising companies. Imagine having a heart attack and when you get home all you see is ads for life insurance.
Giving each person in a hospital a $300 stethoscope + $200 phone + $120 for the subscription + a whole new IT team is simply too expensive when hospitals already struggle to afford basic supplies like vaccines.
I work in a medical environment where the doctors are using AI to assist in note taking. They ask consent at the beginning of the visit and 99% of people agree. They like to do it before going into their office so they can just leave the mic on and nothing gets missed fiddling with equipment. So I hear it as they go by. There’s no hesitation.
Many of these people are highly suspicious of medical care, authority, the government. All kinds of wacky conspiracy theories going around. Yet nobody bats an eye when asked if their intimate conversation can be recorded and verbatim transcribed to be processed on a remote server. Due to high rates of acceptance among doctors and subsequently their patients, it is on track to be integrated into the computer system.
Other facilities are using similar software but not obtaining consent. I don’t know if the doctors are supposed to but lax about it, or maybe you agree to it in some kind of blanket waiver when you obtain care.
I think in a lot of places you’d have a difficult and uphill battle to get a legal finding that the sound of heart beats would be considered confidential. If you got home and had adverts as a result of it, it would have to be because the facility provided your contact info with it, which isn’t what’s at issue here. Also, if you had a heart attack, you will end up being cared for in a cardiology unit. They don’t really need the sound of your heart to infer cardiac issues if you are on a cardiac unit. Can probably just use location data like GPS, call towers and wifi hotspot collection to do so already.
But you are correct that this will never see application in low-resource situations. It’s a pointless idea. If you don’t have someone who is skilled enough to listen to the heart, then you don’t have anyone skilled enough to to treat the problems that could arise. This will be a gizmo for rich worried well people.
I said if it’s done correctly, so with consent etc. ofc.
You are totally wrong though IMO when it comes to costs, check this startup out if you are curious, I don’t know how they collect their data but their aim is to bring cheap heart scans to the world. You might have an IT team (IDK) but not at each hospital, and probably those things could be standalone if extreme cost savings are needed, otherwise you’d like them connected so when there is a potential problem the real doctor can take over.
That start up is extremely silly.
The concept that early detection is always of benefit in long term disease outcomes has recently come under scrutiny and has been discredited or moderated in a number of areas. PSA tests in average risk, asymptomatic people being very prominent. They were found to cause significantly more harm than benefit which is why you don’t see PSAs (ha) telling you check your PSA all the time anymore.
Claiming that any kind of diagnostic imaging will make a dent in a massively common problem is ridiculous. I see no reason to believe that, just a lot of marketing hype. They don’t even say what specific problems they are supposed to be preventing, or treating, or finding, or whatever. “Cardiac disease” is a large category. It’s like saying “skin problems” or “mental health”.
It looks like these people are claiming to be able to perform a cardiac doppler and/or ultrasound? It specifically says it does not interpret it. It produces the image. Generally speaking, that’s the easier part of the job.
I tried to look up what the value of a doppler might be in screening for cardiac disease. Most analysis and guidelines about cardiac screening do not even mention dopplers. Some mention very specific ultrasounds in specific situations. Always along with a bunch of other stuff.
See example - table 3 summary of reccomendations — try to find anything there that has any apparent relationship to this app.
The most general screening for cardiac disease are blood pressure, family history, a risk calculation (like Framingham which you can do online), and sometimes a few simple blood tests.
All in all the claims being made here are ostentatious and unrealistic. This device may have some kind of application but it is not what is claimed here.
