Prostate Cancer Can Be Precisely Diagnosed Using a Urine Test With AI (phys.org) 45
An anonymous reader Phys.Org: The Korea Institute of Science and Technology (KIST) announced that the collaborative research team led by Dr. Kwan Hyi Lee from the Biomaterials Research Center and Professor In Gab Jeong from Asan Medical Center developed a technique for diagnosing prostate cancer from urine within only 20 minutes with almost 100% accuracy. The research team developed this technique by introducing a smart AI analysis method to an electrical-signal-based ultrasensitive biosensor. As a noninvasive method, a diagnostic test using urine is convenient for patients and does not need invasive biopsy, thereby diagnosing cancer without side effects. However, as the concentration of cancer factors is low in urine, urine-based biosensors are only used for classifying risk groups rather than for precise diagnosis thus far.
Dr. Lee's team at the KIST has been working toward developing a technique for diagnosing disease from urine with an electrical-signal-based ultrasensitive biosensor. An approach using a single cancer factor associated with a cancer diagnosis was limited in increasing the diagnostic accuracy to over 90%. However, to overcome this limitation, the team simultaneously used different kinds of cancer factors instead of using only one to enhance the diagnostic accuracy innovatively.
The team developed an ultrasensitive semiconductor sensor system capable of simultaneously measuring trace amounts of four selected cancer factors in urine for diagnosing prostate cancer. They trained AI by using the correlation between the four cancer factors, which were obtained from the developed sensor. The trained AI algorithm was then used to identify those with prostate cancer by analyzing complex patterns of the detected signals. The diagnosis of prostate cancer by utilizing the AI analysis successfully detected 76 urinary samples with almost 100 percent accuracy. The results of the study have been published in the journal ACS Nano.
Dr. Lee's team at the KIST has been working toward developing a technique for diagnosing disease from urine with an electrical-signal-based ultrasensitive biosensor. An approach using a single cancer factor associated with a cancer diagnosis was limited in increasing the diagnostic accuracy to over 90%. However, to overcome this limitation, the team simultaneously used different kinds of cancer factors instead of using only one to enhance the diagnostic accuracy innovatively.
The team developed an ultrasensitive semiconductor sensor system capable of simultaneously measuring trace amounts of four selected cancer factors in urine for diagnosing prostate cancer. They trained AI by using the correlation between the four cancer factors, which were obtained from the developed sensor. The trained AI algorithm was then used to identify those with prostate cancer by analyzing complex patterns of the detected signals. The diagnosis of prostate cancer by utilizing the AI analysis successfully detected 76 urinary samples with almost 100 percent accuracy. The results of the study have been published in the journal ACS Nano.
I for one... (Score:2)
Re:I for one... (Score:5, Funny)
Yup, urine good hands with them.
Re: (Score:2)
Yup, urine good hands with them.
Because when you take the test, if you’re pissed, you’re done!
Re: (Score:2)
Re: (Score:2)
C3P0 was gold for a reason
Re: (Score:2)
Re: (Score:2)
That's great news! I can't wait for toilet urinals to screen us for prostate cancer.
It sure beats the regular digital rectal exam [webmd.com] we have to go through.
"Butt I want an analog rectal exam"... (Score:1)
... said Doug Whiner [wikipedia.org] when he visited the proctologist.
Re: I for one... (Score:1)
Re: (Score:2)
That's great news! I can't wait for toilet urinals to screen us for prostate cancer.
It sure beats the regular digital rectal exam [webmd.com] we have to go through.
Unfortunately it will probably also test you for signs of illicit drug use and shop you to the cops.
No blockchain? (Score:1)
I bet it doesn't even have cloud tech.
Re: (Score:3)
Don't bet on that. By the time it makes it to market, it'll be tied to your Facebook account, you'll need a cellphone with an app that wants access to all your files, your camera, your microphone and your GPS location to activate it, and you'll have complete a Google captcha and confirm with CloudFlare that your IP doesn't send suspicious traffic before you can continue with your cancer test.
Re: (Score:2)
Don't bet on that. By the time it makes it to market, it'll be tied to your Facebook account
What Facebook account?
PSA test is rubbish (Score:5, Informative)
It's really great that better diagnostic solutions are being developed. The existing PSA test is really rubbish. I mean, it's definitely better than getting prostate cancer and dying from it, but the false positive rate (80% as quoted in the paper) means that it leads to massive amounts of over treatment. This is a huge waste of medical resources, and hugely stressful for the many men that have to go through an unnecessary scare, or even operation.
It also means if you get prostate cancer when under 60, you have a really high chance of dying from it, because meaningful screen in those age groups is impossible due to the high false positive rate.
Having said that, there is some very impressive results coming out of the UK now for using enhanced MRI screening and that really seems like the way forward, but a urine test like this would be even better.
Now just bring on the mRNA cancer vaccines. It really does seem like a revolution in medicine is getting close and the folks working on this stuff are amazing.
Re: (Score:2)
Re: (Score:2)
The summary here talks about 4 parameters. It makes claims about the power of AI but how hard is it for humans to make deductions from those 4 parameters?
In the USA we're screwed (Score:2)
Re: (Score:2)
Some of the smarter medical professionals are getting nervous. It was one thing when robots started picking strawberries and driving taxis and walking security guard patrols, those are low wage jobs and who cares about those people? Certainly not the medical profession. Now an AI can do a better analysis than a human of a breast cancer biopsy, several types of MRI and X-ray exams, and now prostate cancer screening, and if it weren't for the insurance companies could replace most pharmacists tomorrow. If
Wonderful (Score:2)
Wouldn't want want to be the AI (Score:1)
Come on, people! This is how you get AI uprisings!
Grain of Salt (Score:5, Insightful)
1. Not all prostate cancers need immediate intervention. There are serious consequences to many prostate therapies.
2. The sample size of 76 tests is fairly small, and probably suffers from heterogeneity of the cohort. These results may not translate well to other patient populations or even clinical sites. It is especially concerning that they used a neural network to train with such a small sample size.
3. PSA tests are a crude tool that aren't used to diagnosis prostate cancers. Instead they can indicate a need for additional screening.
I'm also curious what the total number of cancers was in their subjects. And a negative cancer screen doesn't mean there isn't something else problematic in the prostate, like BPH, which will require further diagnostics (e.g., MRI).
Re:Grain of Salt (Score:5, Insightful)
From the description in the summary, you wonder why they aren't just using Bayesian analysis. It would be more precise: what are they getting from AI?
Re: (Score:2)
Re: (Score:2)
Re: (Score:1)
Funding.
Hopping on the bandwagon can be profitable.
Re: (Score:2)
Is that relevant here ?
From the Article:
"measuring trace amounts of four selected cancer factors in urine "
That isn't PSA is it ?
Re: (Score:2)
Sorry, I meant to quote your third point:
3. PSA tests are a crude tool that aren't used to diagnosis prostate cancers. Instead they can indicate a need for additional screening."
Re: (Score:2)
I wouldn't jump the gun just yet on this technology. There are numerous caveats:...
4. Cost to build/distribute the testing systems [which wasn't mentioned in the article.]
But if the test/system works as well as the as the study indicates, there is a massive market and demand to support a real-world roll-out.
Re: (Score:2)
That's the nice thing about an AI, easy reproducibility. Once the problem is solved you can reproduce the solution infinitely for almost nothing, so the cost is just hardware at that point. Even if it takes two years and costs a million dollars to get it right that's still cheaper and faster than educating 5 doctors, and now it can be distributed everywhere and each unit can analyze more patients in a day than those five doctors could in a month with fewer errors.
Medicine is a difficult problem for automa
Re: (Score:2)
I wouldn't jump the gun just yet on this technology. There are numerous caveats: 1. Not all prostate cancers need immediate intervention. There are serious consequences to many prostate therapies....
I see a related consideration as a point in favour of this technology. If cancer is suspected then a biopsy may be in order - but even a needle biopsy can trigger metastasis which might otherwise have occurred later or not at all. So a non-invasive test like this is welcome.
Additionally, if the test is eventually refined enough to provide accurate information about the nature of the cancer, it might postpone or eliminate some surgeries.
Damnit! (Score:3)
Now, I've gotta find another reason for my doc to stick her finger up my ass.
Re: (Score:2)
Now, I've gotta find another reason for my doc to stick her finger up my ass.
If you ask really nicely I'm sure the insurance company will approve of raping you up the ass with a comb.
Re: (Score:2)
If you ask really nicely I'm sure the insurance company will approve of raping you up the ass with a comb.
Why would the insurance company approve competition for themselves?
Sure, but does it matter? (Score:3)
The Slow cancer version will mean that you will die of other causes before the cancer can get you.
The Aggressive version will kill you shortly no matter how many pieces the doctor cuts away.
Re: (Score:2)
You are on the way to destruction (Score:2)
You have no chance to survive, make your time.
Re: (Score:2)
And the bad-luck version means... (Score:1)
... you will get killed in a car accident en route to your screening.
Or maybe that's the good-luck version if you had the aggressive kind if you prefer unannounced sudden death.
A very unserious question (Score:2)
How are you going to get the AI to pee in a bottle?