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

AI used for skin cancer checks at London hospital

bbc.com

Post reply on HN
Print view View original

61–70 of 91 posts

Re: AI used for skin cancer checks at London hospital

#61

Post by NoPicklez »

andy99 wrote: post #7

99% accuracy in diagnosing benign case This is meaningless. The only thing that matters in this kind of application is false negative rate at some acceptable false positive rate. I assume whoever is working on this knows that, so this is mostly a criticism of the article. That said, this is a horrible use of AI.

I don't think its a horrible use of AI at all, in fact it seems like a fairly good use case.

It's a tool that can be used in amongst the current methods to help detect skin cancer, it shouldn't be used at the only method.

Re: AI used for skin cancer checks at London hospital

#62

Post by QuantumGood »

Scoundreller wrote: post #31
qq66 wrote: post #24

Earlier quoted context omitted.

Sadly, a friend of mine died from this type of problem. He traveled from Ghana to Jordan and fell ill in Jordan, the Jordanian doctors didn't diagnose it as malaria in time to save him because Jordan doesn't have malaria. I'm sure it would have been obvious to a Ghanaian doctor.

Friend did a lot of work in Africa. Got back to the US and starts getting symptomatic and says to himself “I think I got malaria”. Goes to the biggest university hospital nearby he can find. Was initially dismissed but waited it out for the infectious diseases specialist and they quickly agreed with his self-diagnosis. They kept them in hospital for a few days so a parade of clinical students/residents could come by…

I had a weird cardiac issue, for which they kept me in the hospital for a few days and brought in more and more doctors to hear my story. Added tens of thousands to my bill. Eventually checked myself out "against doctor's advice".

Re: AI used for skin cancer checks at London hospital

#64

Post by _kb »

asdefghyk wrote: post #37
ircshotty wrote: post #12

Anecdotally, when I had a potential skin cancer checked at a London hospital they were completely ill-prepared. When I came back to Australia, it was checked and immediately removed as an obvious melanoma. Perhaps the idea of Comparative Advantage also applies to healthcare between countries with natural variances to types of disease?

Australia, Queensland, Brisbane has the highest incidence of skin cancer in World. So Drs and dermatologists would have more experience detecting it here. UV is probably much less intense in England . This link graph shows skin cancer is over 2x in Australia https://www.statista.com/statistics/1032114/countries-with-t...

Queenslander here. There’s also a fairly well equipped imaging centre that’s being testing this form of diagnosis for 5 years: https://acemid.centre.uq.edu.au/research/cre-skin-imaging-an...

Re: AI used for skin cancer checks at London hospital

#65

Post by markdown »

asdefghyk wrote: post #37
ircshotty wrote: post #12

Anecdotally, when I had a potential skin cancer checked at a London hospital they were completely ill-prepared. When I came back to Australia, it was checked and immediately removed as an obvious melanoma. Perhaps the idea of Comparative Advantage also applies to healthcare between countries with natural variances to types of disease?

Australia, Queensland, Brisbane has the highest incidence of skin cancer in World. So Drs and dermatologists would have more experience detecting it here. UV is probably much less intense in England . This link graph shows skin cancer is over 2x in Australia https://www.statista.com/statistics/1032114/countries-with-t...

> Australia, Queensland, Brisbane

OT but what an odd way to address a place. I know it's called down under but not everything there is back to front.

Re: AI used for skin cancer checks at London hospital

#66

Post by Spooky23 »

kennyadam wrote: post #53
Spooky23 wrote: post #33

Earlier quoted context omitted.

It’s really all about the protocol. AI tends to spot things that doctors don’t, and vice versa. Dermatology is also an area being pillaged by private equity and access is poor for many people. I lost my wife to melanoma. She noticed a lesion within days of it appearing, and a doctor saw it within 48 hours and felt it was benign. My wife didn’t accept it and had a plastic surgeon remove it and biopsy, then had a margi…

Similar story: mom had a melanoma removed from her foot. multiple lymph node biopsies and other tests said it had been successfully removed. She went back once a year for checks for 4-5 years. It was only when she was called into see an oncologist and told an unrelated x-ray lead them to discover she had stage 4 metastatic melanoma (brain, liver, spine, femur, lungs and i’m sure i’m forgetting something) that we foun…

My wife had a similar experience, except it was her shoulder. They took a margin, looked at lymph nodes, and did a nuclear test that traced something that I can’t recall.

Another lesson learned is that if at all possible, go to a national cancer center. Even if for a second opinion analysis. The level of care is different and better than what you find in community oncology or hospital practices.

Re: AI used for skin cancer checks at London hospital

#67

Post by nneonneo »

qq66 wrote: post #24
ircshotty wrote: post #12

Anecdotally, when I had a potential skin cancer checked at a London hospital they were completely ill-prepared. When I came back to Australia, it was checked and immediately removed as an obvious melanoma. Perhaps the idea of Comparative Advantage also applies to healthcare between countries with natural variances to types of disease?

Sadly, a friend of mine died from this type of problem. He traveled from Ghana to Jordan and fell ill in Jordan, the Jordanian doctors didn't diagnose it as malaria in time to save him because Jordan doesn't have malaria. I'm sure it would have been obvious to a Ghanaian doctor.

A good family friend caught abdominal TB while abroad, came back to Canada, and was terribly misdiagnosed. TB is very rare here; abdominal TB even more so, and early symptoms look like a lot of other diseases. It took several hospital visits before the doctors realized the actual problem. He very nearly died, and spent nearly a year in the hospital recovering from a parade of complications.

Re: AI used for skin cancer checks at London hospital

#68

Post by TylerE »

1shooner wrote: post #49
qq66 wrote: post #24

Earlier quoted context omitted.

Sadly, a friend of mine died from this type of problem. He traveled from Ghana to Jordan and fell ill in Jordan, the Jordanian doctors didn't diagnose it as malaria in time to save him because Jordan doesn't have malaria. I'm sure it would have been obvious to a Ghanaian doctor.

When I first heard the clinical idiom "When you hear hoof-beats, think horses not zebras" I thought it was a precautionary saying about the bias toward assuming the familiar. But it's meant to be instructive!

No, that’s backwards. It means what it says. When you see a symptom first look for the common things it could be, not the one in a million chance.

Re: AI used for skin cancer checks at London hospital

#69

Post by dijksterhuis »

cmcaleer wrote: post #59
jampekka wrote: post #26

Earlier quoted context omitted.

It's quite easy to correctly classify 100% of benign cases as benign.

If it's so easy, then why do people die from having lesions misdiagnosed as benign? Even if the success rate of the human eye was in the 99.5%+ range, why not have an extra sanity check from an AI model?

> If it's so easy, then why do people die from having lesions misdiagnosed as benign?

You're confusing False Negatives with True Negatives. For Non-Benign (Positive) vs. Benign (Negative) classification:

* True Positive Rate (TPR): non-benign classified as non-benign.

* False Positive Rate (FPR): benign misclassified as non-benign.

* True Negative Rate (TNR): benign classified as benign.

* False Negative Rate (FNR): non-benign misclassified as benign.

> It's quite easy to correctly classify 100% of benign cases as benign.

You can engineer a 100% TNR if you just classify everything as the "benign" negative class. The FNR is going to be 100% too, but that doesn't matter -- you correctly classified 100% of benign cases as benign.

> why do people die from having lesions misdiagnosed as benign?

Because the FNR is not 0%. FNR is important. You probably want a decent TPR in there as well. And FPR can be very important too, depending on how life-changing/painful/invasive the treatment for a positive case is!

Re: AI used for skin cancer checks at London hospital

#70

Post by amarcheschi »

floppydiscen wrote: post #5

This would be such a beautiful use of AI

I'm attending a course with some goat professors, that aim to teach this kind of things. Chances are, given it's a small course you won't produce anything of real value, but the aim should be to analyze 3 datasets from 3 different neurodegenerative diseases, find some common markers and eventually make a ml model to better diagnose people with these diseases. It's not exactly easy, but given it's a real world problem it's extremely motivating

(we've been also allowed to use a university vm with 3tb of ram and that's nice)

Post reply on HN