Microsoft and Paige Are Building the World's Largest AI Model for Detecting Cancer (cnbc.com) 7
Microsoft is teaming up with digital pathology provider Paige to build the world's largest image-based artificial intelligence model for identifying cancer. From a report: The AI model is training on an unprecedented amount of data that includes billions of images, according to a release. It can identify both common cancers and rare cancers that are notoriously difficult to diagnose, and researchers hope it will eventually help doctors who are struggling to contend with staffing shortages and growing caseloads. Paige develops digital and AI-powered solutions for pathologists, which are doctors who carry out lab tests on bodily fluids and tissues to make a diagnosis. It's a specialty that often operates behind the scenes, and it's crucial for determining a patient's path forward.
"You don't have cancer until the pathologist says so. That's the critical step in the whole medical edifice," Thomas Fuchs, co-founder and chief scientist at Paige, told CNBC in an interview. But despite pathologists' essential role in medicine, Fuchs said their workflow has not changed much in the last 150 years. To diagnose cancer, for instance, pathologists usually examine a piece of tissue on a glass slide under a microscope. The method is tried and true, but if pathologists miss something, it can have dire consequences for patients.
"You don't have cancer until the pathologist says so. That's the critical step in the whole medical edifice," Thomas Fuchs, co-founder and chief scientist at Paige, told CNBC in an interview. But despite pathologists' essential role in medicine, Fuchs said their workflow has not changed much in the last 150 years. To diagnose cancer, for instance, pathologists usually examine a piece of tissue on a glass slide under a microscope. The method is tried and true, but if pathologists miss something, it can have dire consequences for patients.
Where is Watson? (Score:2)
I was made to believe that Watson could already do that a decade ago.
Re: (Score:2)
IBM Watson is like old version of ChatGPT. It doesn't look images.
But Google Deepmind has been looking images and detecting cancer.
"The research says the AI model could predict breast cancer with the same level of accuracy as a single expert radiologist. Compared to human experts, the system saw a reduction in false positives by 5.7 per cent in the US and 1.2 per cent in the UK, and in false negatives of 9.4 per cent in the US and 2.7 per cent in the UK."
https://www.wired.co.uk/articl... [wired.co.uk]
Image Data? (Score:1)
Re: (Score:2)
Sharing medical information has always been crucial to education, research, writing papers, and patient care. The stipulation is that there can be no identifiable patient information. If I submit a photo or write a case study or give a presentation, I can say "This is a 37 year old man with cirrhosis" or ""this x-ray is from a 59 year old woman with a history of valley fever and a new lung mass". Nothing in that identifies a specific person, nor could identity ever be inferred as those are common conditi
Re: (Score:1)
Very questionable that anonymized patient data cannot pinpoint the true identity. You're providing one side of the story and its apparent which one.
https://hardware.slashdot.org/... [slashdot.org]
"Why Anonymized Data Isn't"
Ars has a review of recent research, and a summary of the history, in the field of reidentification — identifying people from anonymized data. Paul Ohm's recent paper is an elaboration of what Ohm terms a central reality of data collection: "Data can either be useful or perfectly anonymous but never both." ... For almost every person on earth, there is at least one fact about them stored in a computer database that an adversary could use to blackmail, discriminate against, harass, or steal the identity of him or her. I mean more than mere embarrassment or inconvenience; I mean legally cognizable harm. ... Reidentification science disrupts the privacy policy landscape by undermining the faith that we have placed in anonymization."
"...in 2000, [researcher Latanya Sweeney] showed that 87 percent of all Americans could be uniquely identified using only three bits of information: ZIP code, birthdate, and sex.
Re: (Score:1)
There is literally a whole field in data science, reidentification. What is your motivation for telling people they can't be reidentified?
You don't have cancer till the pathologist says so (Score:2)