Computer better than doctor at breast cancer diagnosis

pink ribbonSCA Svenska Cellulosa Aktiebolaget

Stanford University computer scientist Dr. Daphne Koller may have created an electronic pathologist. Software can now grade breast cancer cells as "indolent" or "aggressive" more accurately than a human.

Called C-Path (Computational Pathologist), the software never gets tired, never relaxes, and even identifies structures that were unknown in the past to more accurately define the types of breast cancer cells. First defining 6,642 items to look for in breast-cancer biopsies, Dr. Koller then let the software learn what was important. The software also predicts how the patient will fare five years after treatment.

Interestingly, of the 11 features most useful for the software, three were non-cancerous stroma cells, not always considered by pathologists. They play more of a role in supporting tumors than previously thought. Thanks, intelligent software.

Amazing

Another interesting point is that this technology could more accurately give someone life-expectancy results.

end2760 on economist.com

One thing the article doesn't point out is that AI methods in the past did a poor job of detecting breast cancer. This is an important new result and big step for the field.

kenjackson on news.ycombinator.com

This technology is already amazing and its completely new. I can only imagine after modifications and more technological advances in the next years how far it could truly take us.

bdevin7 on economist.com

Still need people

This technology will save lives, but it will also put people in various positions out of jobs. I feel that in the coming years we as a society are going to have to deal with these issues and sort out which options are best for the majority.

djavery on economist.com

But your statement "Beyond the immediate implications — hospitals around the world should get their hands on C-Path ASAP" implies you didn't actually read the the paper yourself - because they can't.

Michael Linde on extremetech.com

The funny thing is, most rads physicians would probably welcome this, as mammography is often the least preferred sub-specialty (my wife is a radiologist and feels this way, and I've encountered many radiologists with the same outlook).

separated on news.ycombinator.com

Still need some more help

But the real game changer in the cancer diagnosis and treatment field will be through nanobiotechnology.

robertbaxter on economist.com

Now the bad news is that human medicine moves super slowly due to very strict regulations, and I expect doctors to resist this. I hate to say this, but I think health insurance companies are our best hope to push this technology into greater use.

bh42222 on news.ycombinator.com

Every patient is different, and only humans are able to take into account all of the free and unique parameters in a cancer diagnosis. Many things, such as whether a patient will survive, involve factors which would confound a computer.

Realities on economist.com

Just two days ago we talked about how AI may take our jobs, and here's another example. Are you scared at the job loss potential, or thrilled that smarter computers will help is in many new ways?

Insider: How the basic tech behind the Internet works
Join the discussion
Be the first to comment on this article. Our Commenting Policies