Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Towards accurate differential diagnosis with large language models

arxiv.org

Post reply on HN
Print view View original

91–99 of 99 posts

Re: Towards accurate differential diagnosis with large language models

#91

Post by corethree »

rscho wrote: post #85
corethree wrote: post #80

Earlier quoted context omitted.

Are you a US doctor? I urge you to explain why you charge so much and have the poorest performance out of all 1st world countries? It may change your perspective on treating patients like idiots when really the US doctor is the epitome of fraudulent incompetence. Over testing is a problem. The way to lessen over testing is to not be an incompetent doctor so that patients don't take things into their own hands. Barrin…

I have worked in both the US and Europe. I am back in Europe now, as I vastly prefer it over the US. > Even if the disease is extremely unlikely, if the possibility remains valid based on the given symptoms and the patient requests the test you should allow it. Precisely not. That's exactly how you get unusable results with super expensive healthcare, because the follow up that a positive test dictates is often worse…

Also, the other part of this that's fishy is why can't the test results be reported to the patient? Your reasoning makes sense for over testing. But does it make sense to censor a doctor prescribed test from the patient? No it doesn't. The law is made to make the doctor central to treatment. The patient must schedule another appointment and must go through the doctor in order to treat himself.

There's too much business incentive here. You label me as toxic but there's a reason why patients in the US have such a vicious attitude toward docs.

Re: Towards accurate differential diagnosis with large language models

#92

Post by rscho »

corethree wrote: post #91
rscho wrote: post #85

Earlier quoted context omitted.

I have worked in both the US and Europe. I am back in Europe now, as I vastly prefer it over the US. > Even if the disease is extremely unlikely, if the possibility remains valid based on the given symptoms and the patient requests the test you should allow it. Precisely not. That's exactly how you get unusable results with super expensive healthcare, because the follow up that a positive test dictates is often worse…

Also, the other part of this that's fishy is why can't the test results be reported to the patient? Your reasoning makes sense for over testing. But does it make sense to censor a doctor prescribed test from the patient? No it doesn't. The law is made to make the doctor central to treatment. The patient must schedule another appointment and must go through the doctor in order to treat himself. There's too much busine…

Many people feel the way you do, and it's understandable. Those laws vary from country to country, but variations of the same are found throughout western countries. It's in part a remain of the paternalistic medicine of olden times, where docs were supposed to decide for patients. But the main goal today is to avoid patients learning about life-changing news (AIDS, cancer, etc.) without being supported emotionally and informed about what results imply, to avoid rash or violent reactions (suicide) due to misunderstandings. I personally experienced grave misunderstandings over news that seemed trivial to me, but that the patient took as extremely aggravating. Of course this rule is very profitable as well for private care providers (which I am not, BTW). At this point, whether or not the rule is justified becomes a political issue, and I don't think there's a solution that would satisfy everybody.

Re: Towards accurate differential diagnosis with large language models

#93

Post by corethree »

rscho wrote: post #92
corethree wrote: post #91

Earlier quoted context omitted.

Also, the other part of this that's fishy is why can't the test results be reported to the patient? Your reasoning makes sense for over testing. But does it make sense to censor a doctor prescribed test from the patient? No it doesn't. The law is made to make the doctor central to treatment. The patient must schedule another appointment and must go through the doctor in order to treat himself. There's too much busine…

Many people feel the way you do, and it's understandable. Those laws vary from country to country, but variations of the same are found throughout western countries. It's in part a remain of the paternalistic medicine of olden times, where docs were supposed to decide for patients. But the main goal today is to avoid patients learning about life-changing news (AIDS, cancer, etc.) without being supported emotionally a…

Well how about letting the patient choose how its reported. And even letting the patient be aware that the results might be life threatening. This will satisfy everybody.

Most tests aren't for things like testing for HIV. So your reasoning makes zero sense as this law is applied to every test. And the whole suicide thing is obviously a miniscule issue. You're just making that crap up to sow controversy. Yeah I'm sure you encountered dozens of patients ready to kill themselves over test results but only your kind words were able to completely assuage them and stop them... Yeah sounds like a load of bullshit to me.

There's no controversy here. It's a business incentive through and through.

Patient choice would satisfy all patients. Only doctors wouldn't be satisfied. And that's the problem.

Patients are dealing with there lives here. And Even you as a third party are painting the situation as "controversy" because you've been so desensitized by the patients you're supposed to take care of.

Even though you're not in private care you are part of the problem because you're blind to the problem. I know multiple patients who had surgery in America only to go to another country for a check up for the doctor in the other country to say the surgery was completely pointless.

You know how serious that is? I don't think you get it. Painting a crime as "controversial" or "political" issue is extremely aggravating for the person who's life is at stake. The only reason you characterize it this way is because you're not the one who's life is at stake.

Patients aren't toxic. Patients are desperate. And most doctors including you just don't give enough of a shit. I'm sorry but that's the truth.

Re: Towards accurate differential diagnosis with large language models

#94

Post by chrstfer »

ugh123 wrote: post #10
guender wrote: post #2

Yikes, do I read this correctly and the LLM alone outperforms clinician + LLM?

It also outperforms clinician + googling which the study observed (many doctors use google search to help with diagnosis, terminology, recent studies, etc)

Hell, the LLMs themselves outperformed clinician plus LLM...

Re: Towards accurate differential diagnosis with large language models

#95

Post by chrstfer »

twobitshifter wrote: post #88
guender wrote: post #2

Yikes, do I read this correctly and the LLM alone outperforms clinician + LLM?

Makes sense. In many applications the ‘human in the loop’ will become the weakest link. Imagine giving alpha go to players as a suggested move assistant. Even the best players would have seen alpha go’s suggestions and thought they were an error. Can we eventually trust AI even if we can’t understand its reasoning?

They'll be able to explain their reasoning......

Re: Towards accurate differential diagnosis with large language models

#96

Post by rscho »

corethree wrote: post #93
rscho wrote: post #92

Earlier quoted context omitted.

Many people feel the way you do, and it's understandable. Those laws vary from country to country, but variations of the same are found throughout western countries. It's in part a remain of the paternalistic medicine of olden times, where docs were supposed to decide for patients. But the main goal today is to avoid patients learning about life-changing news (AIDS, cancer, etc.) without being supported emotionally a…

Well how about letting the patient choose how its reported. And even letting the patient be aware that the results might be life threatening. This will satisfy everybody. Most tests aren't for things like testing for HIV. So your reasoning makes zero sense as this law is applied to every test. And the whole suicide thing is obviously a miniscule issue. You're just making that crap up to sow controversy. Yeah I'm sure…

For the record, I'm not in favor of such rules. Yes, you're really toxic, and display the typical US reaction. I'm fucking relieved that I don't have to deal with US egos and all this US political shit anymore. Peace out, and take care.

Re: Towards accurate differential diagnosis with large language models

#97

Post by twobitshifter »

chrstfer wrote: post #95
twobitshifter wrote: post #88

Earlier quoted context omitted.

Makes sense. In many applications the ‘human in the loop’ will become the weakest link. Imagine giving alpha go to players as a suggested move assistant. Even the best players would have seen alpha go’s suggestions and thought they were an error. Can we eventually trust AI even if we can’t understand its reasoning?

They'll be able to explain their reasoning......

Explaining and understanding the explanation are different.

Re: Towards accurate differential diagnosis with large language models

#98

Post by corethree »

rscho wrote: post #96
corethree wrote: post #93

Earlier quoted context omitted.

Well how about letting the patient choose how its reported. And even letting the patient be aware that the results might be life threatening. This will satisfy everybody. Most tests aren't for things like testing for HIV. So your reasoning makes zero sense as this law is applied to every test. And the whole suicide thing is obviously a miniscule issue. You're just making that crap up to sow controversy. Yeah I'm sure…

For the record, I'm not in favor of such rules. Yes, you're really toxic, and display the typical US reaction. I'm fucking relieved that I don't have to deal with US egos and all this US political shit anymore. Peace out, and take care.

Like every other patient my life and my money are more important than anything else. It has nothing to do with ego.

I can't respect a doctor who doesn't understand that and thinks it's just toxic. You're lucky to be practicing Europe.

Re: Towards accurate differential diagnosis with large language models

#99

Post by kelipso »

bradknowles wrote: post #58
corethree wrote: post #57

Earlier quoted context omitted.

You don't know how these doctors view you. They view you as a "non-expert" as one commenter put it. Basically you're stupid to them and your thoughts are pointless. The only way to convince them is to sue. But there's nothing worth suing here for, you got the test. You can only sue if damage was done, then the doctor would be responsible for blocking the test when you explicitly requested it.

I need to find a doctor who can be a more active partner in helping me diagnose and address my problems. If you have any suggestions in the Austin area, please let me know.

Best bet would be make a lot of money and get a concierge doctor haha. Otherwise you're what, probably a 10 minute a month appointment to them and whatever happens to you, they can add it to their statistics.
Post reply on HN