6 ms·

What about the 6%. Are they better - let's use those!

by nikita 5y ago

maweki 5y ago
We just need worse radiologists.
- jeffbee 5y ago
  Under-rated comment here. The question is not whether some AI beats experts, even though the conclusion of this paper is clearly that the state-of-the-art AI does in fact beat them, but whether they are going to beat the radiologist at the free clinic in some poorly-served place.
  - nradov 5y ago
    The US healthcare system has little correlation between price and care quality. More expensive providers don't reliably delivery better patient outcomes.
    - wolverine876 5y ago
      Do you know where there is more info on that?
      - nradov 5y ago
        This is a good start. https://journalofethics.ama-assn.org/article/complex-relationship-between-cost-and-quality-us-health-care/2014-02 https://journalofethics.ama-assn.org/article/complex-relatio... https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4863949/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4863949/
- 256lie 5y ago
  Even the average radiologist is high variable, not to mention inter-reader variability.
acchow 5y ago
I agree. I think the implied conclusion that “none of the AI know what they’re doing!” is disingenuous. This is like the comparisons of 200 top hedge funds to index performance and concluding that hedge funds can’t beat the market. Ignoring the other conclusion: most hedge funds suck Edit: or most hedge funds are fraudulent?
- eklbt 5y ago
  This hedge fund example is not great because the hedge funds that beat the market(and overcome the added management fees) can't do so consistently.
  - qweqwweqwe-90i 5y ago
    But many do... e.g. rentec
    - sidlls 5y ago
      Statistically it stretches credulity to the breaking point for an outlier be an outlier year after year, whether it's in this or any other area.
  - spywaregorilla 5y ago
    "Hedge funds" generally aren't supposed to beat the market. They're supposed to do different to the market. That might be, for example, generally a bit worse but hugely better if the market suffers a catastrophic loss for a simple example.
- aeturnum 5y ago
  > This is like the comparisons of 200 top hedge funds to index performance and concluding that hedge funds can’t beat the market. I think you've misread what conclusion you're supposed to draw from this kind of assessment. For both the hedge funds and this the conclusion isn't that the job is impossible - but that we lack the tools to predict who will do the job. You can't know which hedge fund will outperform (though some will) and we don't seem to be able to predict which AI models will perform well enough to use. Also that your odds of picking a winner "from the field" are quite bad. It doesn't mean there's anything wrong with using a hedge fund - but there's a level of risk that one shouldn't ignore. Like, my conclusion from this is that I would not accept a pure AI solution (because most of them are bad), but I would be interested in AI assistance. For many people, the promise is in replacing not supplementing doctors and this is the same as failure.
Retric 5y ago
In the test they scored better than one radiologist, but worse than two working together. However: https://xkcd.com/882/ https://xkcd.com/882/
- bpodgursky 5y ago
  I've never personally had a radiology scan where two radiologists worked together on it.
  - cratermoon 5y ago
    That just means the scans were easy to interpret and/or the consequences of error were not mortal. While that's most scans, the ones where we care about the most accurate results are the ones the AI is worst.
  - pvarangot 5y ago
    I never had anything serious, just sebaceous cysts and fascia inflammation. On both of those two radiologists worked on them, and everyone was pretty sure it was not cancerous or anything. I think two radiologists are standard for most things where they may recommend a biopsy. I asked about AI and they couldn't use the one they had on the cyst because it was not trained for ankles.
bpodgursky 5y ago
Yeah... I feel like a lot of people miss that in only takes one working implementation to change everything. Imagine like "94% of all attempts at heavier-than-air flight crash on first attempt". Sure... kinda misses the point.
- ChemSpider 5y ago
  Recent example: 99% of all software to solve protein structures is crap. And then there is AlphaFold.
- pessimizer 5y ago
  Are you claiming that they found two AI systems that beat a skilled radiologist consistently? Because if they had, that would be the headline. People aren't stupid.
ChemSpider 5y ago
Exactly. PLUS: I am sure the radiologists used for this paper paid extra attention to their diagnosis, because they knew it was a competition and colleagues are watching. I am sure the average radiologist, tired and bored at 8am in the morning, is much worse.
- mandevil 5y ago
  This is why the standard practice today for detecting cancer is to use two radiologists to look at every image- to account for that sort of variability. And you can see that a small number of AI's were better than a single doctor, because of that human variability and frailty. None of them were better than two, however, because that variability can be adjusted for, and computers can't. Maybe they will, eventually, but I'm much less optimistic than I was <mumble mumble> years ago when I was 23 and working on vehicle autonomy.
somenewaccount1 5y ago
0% of the studies were better than 2 radiologists reviews - which is the current practice. The 6% only scored better when compared to a single radiologist review. That's why we use double reviews.
- TameAntelope 5y ago
  What about 1 AI and 1 radiologist, then? Double the capacity for reviews.
  - kenjackson 5y ago
    Or the consensus of two AIs.
    - sciclaw 5y ago
      Yes, but when disagreements occur, it's hard to make AI systems "talk" and convince one another
- lavishlatern 5y ago
  > That's why we use double reviews. We being Europeans. Double reads are not standard in the majority of the world, including the US.
avereveard 5y ago
selecting survivors of comparative study without understanding where the performances come from gives you no guarantee of future performances - which is more or less the issue with all these complex multivariate regressions that journalist, startups and venture capitalists insist calling AIs.
striking 5y ago
The full snippet editorialized by the post is > Thirty four (94%) of 36 AI systems evaluated in these studies were less accurate than a single radiologist, and all were less accurate than consensus of two or more radiologists.
- cratermoon 5y ago
  Blame the character limit on HN titles. How would you would have written it to fit in the limit?
  - OJFord 5y ago
    Not GP, but I'd prefer '34' over '94%'. (Especially since the latter was parenthesised - I'd remove the parenthetical before anything else.) (Or just 94%, and then have people say 'but N= only 36!'.) Percentages for small numbers, in particular below 100, are just annoying and often designed to mislead. (Though I'm not suggesting that here, it's standard in medicine if not other fields.)
  - yawnxyz 5y ago
    My crack at this (only for fun! Your title is great): "Only 1 of 36 AI systems was more accurate than a single radiologist"
  - striking 5y ago
    The guidelines say you should use the title as-is, but my attempt might put the focus on the "and 100% were worse than 2 radiologists" part, to mitigate exactly the problem of "well let's use the 6% that actually work" comments that exist without the context.
256lie 5y ago
The article is simplified (a retrospective metastudy) and might not be indicative of what real-life performance. Even reader studies (which would be more rigorous) skip so much that would be crucial to actual deployment (integration into the clinical workflow being one such critical factor).
sciclaw 5y ago
I'm happy someone made a comment like this. I work an an AI company where we screen for Diabetic Retinopathy. The company is a decade old, and has validated in huge studies (over 100k patients). It's a hard problem that we have all worked very hard to solve it. At the same time, it's easy to build an AI tool that looks at an image and says healthy or unheathly (or poor quality image). So the bar is low for making something that is appears functional. But hardly anyone makes good AI tools, so studies that look at different AIs systems see lots of the bad ones. It's always a bummer when the headlines are all dismissive of AI in general. Googling "diabetic retinopathy AI" the top result is "Artificial Intelligence Falls Short in Detecting Diabetic Eye Disease" (https://healthitanalytics.com/news/artificial-intelligence-falls-short-in-detecting-diabetic-eye-disease https://healthitanalytics.com/news/artificial-intelligence-f...), yet if you read the article it says one tool is better than humans, which to me, is the real takeaway.
dang 5y ago
We've since changed the title. The submitted title ("94% of 36 AI systems evaluated were less accurate than a single radiologist") broke the site guidelines by editorializing. Submitters: please don't do that. "Please use the original title, unless it is misleading or linkbait; don't editorialize." - https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html

What about the 6%. Are they better - let's use those!

by nikita 5y ago

What about the 6%. Are they better - let's use those!

maweki 5y ago

We just need worse radiologists.

jeffbee 5y ago
Under-rated comment here. The question is not whether some AI beats experts, even though the conclusion of this paper is clearly that the state-of-the-art AI does in fact beat them, but whether they are going to beat the radiologist at the free clinic in some poorly-served place.
- nradov 5y ago
  The US healthcare system has little correlation between price and care quality. More expensive providers don't reliably delivery better patient outcomes.
  - wolverine876 5y ago
    Do you know where there is more info on that?
    - nradov 5y ago
      This is a good start. https://journalofethics.ama-assn.org/article/complex-relationship-between-cost-and-quality-us-health-care/2014-02 https://journalofethics.ama-assn.org/article/complex-relatio... https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4863949/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4863949/
256lie 5y ago
Even the average radiologist is high variable, not to mention inter-reader variability.

acchow 5y ago

I agree. I think the implied conclusion that “none of the AI know what they’re doing!” is disingenuous. This is like the comparisons of 200 top hedge funds to index performance and concluding that hedge funds can’t beat the market. Ignoring the other conclusion: most hedge funds suck Edit: or most hedge funds are fraudulent?

eklbt 5y ago
This hedge fund example is not great because the hedge funds that beat the market(and overcome the added management fees) can't do so consistently.
- qweqwweqwe-90i 5y ago
  But many do... e.g. rentec
  - sidlls 5y ago
    Statistically it stretches credulity to the breaking point for an outlier be an outlier year after year, whether it's in this or any other area.
- spywaregorilla 5y ago
  "Hedge funds" generally aren't supposed to beat the market. They're supposed to do different to the market. That might be, for example, generally a bit worse but hugely better if the market suffers a catastrophic loss for a simple example.
aeturnum 5y ago
> This is like the comparisons of 200 top hedge funds to index performance and concluding that hedge funds can’t beat the market. I think you've misread what conclusion you're supposed to draw from this kind of assessment. For both the hedge funds and this the conclusion isn't that the job is impossible - but that we lack the tools to predict who will do the job. You can't know which hedge fund will outperform (though some will) and we don't seem to be able to predict which AI models will perform well enough to use. Also that your odds of picking a winner "from the field" are quite bad. It doesn't mean there's anything wrong with using a hedge fund - but there's a level of risk that one shouldn't ignore. Like, my conclusion from this is that I would not accept a pure AI solution (because most of them are bad), but I would be interested in AI assistance. For many people, the promise is in replacing not supplementing doctors and this is the same as failure.

Retric 5y ago

In the test they scored better than one radiologist, but worse than two working together. However: https://xkcd.com/882/ https://xkcd.com/882/

bpodgursky 5y ago
I've never personally had a radiology scan where two radiologists worked together on it.
- cratermoon 5y ago
  That just means the scans were easy to interpret and/or the consequences of error were not mortal. While that's most scans, the ones where we care about the most accurate results are the ones the AI is worst.
- pvarangot 5y ago
  I never had anything serious, just sebaceous cysts and fascia inflammation. On both of those two radiologists worked on them, and everyone was pretty sure it was not cancerous or anything. I think two radiologists are standard for most things where they may recommend a biopsy. I asked about AI and they couldn't use the one they had on the cyst because it was not trained for ankles.

bpodgursky 5y ago

Yeah... I feel like a lot of people miss that in only takes one working implementation to change everything. Imagine like "94% of all attempts at heavier-than-air flight crash on first attempt". Sure... kinda misses the point.

ChemSpider 5y ago
Recent example: 99% of all software to solve protein structures is crap. And then there is AlphaFold.
pessimizer 5y ago
Are you claiming that they found two AI systems that beat a skilled radiologist consistently? Because if they had, that would be the headline. People aren't stupid.

ChemSpider 5y ago

Exactly. PLUS: I am sure the radiologists used for this paper paid extra attention to their diagnosis, because they knew it was a competition and colleagues are watching. I am sure the average radiologist, tired and bored at 8am in the morning, is much worse.

mandevil 5y ago
This is why the standard practice today for detecting cancer is to use two radiologists to look at every image- to account for that sort of variability. And you can see that a small number of AI's were better than a single doctor, because of that human variability and frailty. None of them were better than two, however, because that variability can be adjusted for, and computers can't. Maybe they will, eventually, but I'm much less optimistic than I was <mumble mumble> years ago when I was 23 and working on vehicle autonomy.

somenewaccount1 5y ago

0% of the studies were better than 2 radiologists reviews - which is the current practice. The 6% only scored better when compared to a single radiologist review. That's why we use double reviews.

TameAntelope 5y ago
What about 1 AI and 1 radiologist, then? Double the capacity for reviews.
- kenjackson 5y ago
  Or the consensus of two AIs.
  - sciclaw 5y ago
    Yes, but when disagreements occur, it's hard to make AI systems "talk" and convince one another
lavishlatern 5y ago
> That's why we use double reviews. We being Europeans. Double reads are not standard in the majority of the world, including the US.

avereveard 5y ago

selecting survivors of comparative study without understanding where the performances come from gives you no guarantee of future performances - which is more or less the issue with all these complex multivariate regressions that journalist, startups and venture capitalists insist calling AIs.

striking 5y ago

The full snippet editorialized by the post is > Thirty four (94%) of 36 AI systems evaluated in these studies were less accurate than a single radiologist, and all were less accurate than consensus of two or more radiologists.

cratermoon 5y ago
Blame the character limit on HN titles. How would you would have written it to fit in the limit?
- OJFord 5y ago
  Not GP, but I'd prefer '34' over '94%'. (Especially since the latter was parenthesised - I'd remove the parenthetical before anything else.) (Or just 94%, and then have people say 'but N= only 36!'.) Percentages for small numbers, in particular below 100, are just annoying and often designed to mislead. (Though I'm not suggesting that here, it's standard in medicine if not other fields.)
- yawnxyz 5y ago
  My crack at this (only for fun! Your title is great): "Only 1 of 36 AI systems was more accurate than a single radiologist"
- striking 5y ago
  The guidelines say you should use the title as-is, but my attempt might put the focus on the "and 100% were worse than 2 radiologists" part, to mitigate exactly the problem of "well let's use the 6% that actually work" comments that exist without the context.

256lie 5y ago

The article is simplified (a retrospective metastudy) and might not be indicative of what real-life performance. Even reader studies (which would be more rigorous) skip so much that would be crucial to actual deployment (integration into the clinical workflow being one such critical factor).

sciclaw 5y ago

I'm happy someone made a comment like this. I work an an AI company where we screen for Diabetic Retinopathy. The company is a decade old, and has validated in huge studies (over 100k patients). It's a hard problem that we have all worked very hard to solve it. At the same time, it's easy to build an AI tool that looks at an image and says healthy or unheathly (or poor quality image). So the bar is low for making something that is appears functional. But hardly anyone makes good AI tools, so studies that look at different AIs systems see lots of the bad ones. It's always a bummer when the headlines are all dismissive of AI in general. Googling "diabetic retinopathy AI" the top result is "Artificial Intelligence Falls Short in Detecting Diabetic Eye Disease" (https://healthitanalytics.com/news/artificial-intelligence-falls-short-in-detecting-diabetic-eye-disease https://healthitanalytics.com/news/artificial-intelligence-f...), yet if you read the article it says one tool is better than humans, which to me, is the real takeaway.

dang 5y ago

We've since changed the title. The submitted title ("94% of 36 AI systems evaluated were less accurate than a single radiologist") broke the site guidelines by editorializing. Submitters: please don't do that. "Please use the original title, unless it is misleading or linkbait; don't editorialize." - https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html