Facebook Wants To Use Machine Learning To Make MRIs Faster 67
Facebook believes they can use machine learning to speed up magnetic resonance imaging (MRI) scans. Computer scientists from the social networking site are working with New York University's medical school on the project. CNNMoney reports: NYU is providing an anonymous dataset of 10,000 MRI exams, a trove that will include as many as three million images of knees, brains and livers. Researchers will use the data to train an algorithm, using a method called deep learning, to recognize the arrangement of bones, muscles, ligaments, and other things that make up the human body. Building this knowledge into the software that powers an MRI machine will allow the AI to create a portion of the image, saving time. Making the tests faster would allow radiologists to perform a wider variety of tests.
It will probably (Score:2, Funny)
With RFID at the Mandalay Bay, it could be trivial (Score:2)
While the use of a "legacy system" is a bad thing, the NFC tags inside the badges could be easily read. One vendor offered a free RFID wallet for passersby, who yes, had their badges scanned.
Unusually, the name tag didn't have embedded NFC, rather, an additional tag was used. Remove the tag, and no NFC read.
But the UBM contractor who screwed up.... is a Black Spot on their event.
Re: (Score:1)
Wrong story?
Re: (Score:2)
Wrong story.
Re: (Score:2)
moron :-)
Re: (Score:2)
Don't worry. We've all done it. It's all these damn browser tabs the kids have these days. Confusing as hell.
What could possibly go wrong? (Score:2)
Re:What could possibly go wrong? (Score:4, Insightful)
No, this idea is dead on arrival.
As a patient, if I discovered my doctor was using a "best guess" image, which let's face it, that is what this is, I would transfer hospitals instantly.
Re:What could possibly go wrong? (Score:4, Informative)
No, this idea is dead on arrival.
As a patient, if I discovered my doctor was using a "best guess" image, which let's face it, that is what this is, I would transfer hospitals instantly.
Having had a brain MRI a couple of months ago to rule out a tumor causing the nerve problem with my right eye, I completely agree with you. The Ophthalmologist I went to first set it up, and the imaging people then reported no tumor or other problems found. My Ophthalmologist then sent me to see a Neuro-Ophthalmologist at Jefferson University Hospital, where they in turn did their own evaluation of the image, just to make absolutely sure that it had been read correctly the first time. While the eye problem (right eye outside muscle isn't working, there's a blood clot in the vein that feeds the nerve) is annoying with a turned inwards right eye, it's really nice to know that I don't have any tumors or other bad problems in my brain.
Re: (Score:2)
Something as small as that clot could easily fit within a section of the image this AI approach skips scanning and interpolates in a problem-free vein.
Yeah, that thought occurred to me. The vein is tiny, and the clot even smaller. The N-O said that he spent quite a bit of time examining the area around my eye nerves, and the visual cortex in the brain. Another thing that occurred to me is overlooking a brain aneurysm. Had a co-worker fall dead out of his chair at work one day because of that. His father had died of the same thing, and were it in my family, I'd be pestering the doctors for yearly scans to make sure I wasn't developing one.
My own father
Re: (Score:2)
Re: (Score:1)
Re:What could possibly go wrong? (Score:5, Insightful)
Re: (Score:3)
Yeah, I want to know what their ulterior motives are in this enterprise. I don't buy it being solely for "opportunities to license AI software to hospitals" for a second.
Re: (Score:2)
What could possibly go wrong... https://www.theregister.co.uk/... [theregister.co.uk]
Re: (Score:2)
Nvidia were using neural networks to help denoise the use of monte-carlo methods to implement global illumination methods in ray-tracing. Normally, the resulting image will look a bit grainy like classic film movies. But using the DNN, they get to keep texture detail while removing the noise.
They've had similar successes with modelling CFD. Using DNN they were able to get vast speedups while improving accuracy over classic Navier-Stokes equations. That suggests there is some other mathematical model that sh
Re: (Score:2)
hipaa says no! (Score:3)
hipaa says no!
Re: (Score:3)
hipaa says that the information can't identify you.
Read through all those pages you signed before your MRI to see if 'anonymized data' could be used or sold to 3rd parties.
Re: (Score:2)
Really? You think that something that gives an EXACT profile of unchanging physical attributes (your skull for many) cannot be identifiable, and can in any possible way be anonymized?
Interesting interpretation.
Re: (Score:2)
HIPAA indeed says that, but the fact is that you can. There are a number of papers that demonstrate this, you can use an open source program like Horos or FSL to reconstruct your skull and due to the way algorithms work (offsets and distances between body parts) get matches in a Facebook algorithm.
Modern 64-ch 3T MRI has even higher resolution than the papers used (which were 8 and 12 channel 1.5 and 3T). Some of the voxel sizes we get with custom sequences allow you to even (very faintly, but sufficient f
PLEASE REFRAIN FROM USING THE WORD VAGINA (Score:1)
Vagina, pussy, snatch, etc are overly gender specific. Please use the term "front hole".
Sincerely,
State of California
This is a stupid idea (Score:1)
Make sure to read the fine print... (Score:5, Funny)
Re: (Score:3)
Re: (Score:2)
By using the Facebook MRI Scanning Technology, you agree that your MRI scan will be posted to your Timeline once the scan is complete.
Got cancer :(
[37 people liked this]
Core business (Score:1)
Medical Images and AI are old teammates (Score:3)
Anytime anyone wants to train an AI system its always medical images. I did this as my undergrad final year project almost 20 years back. Sure the algos are probably better now but there is something about Medical Images which makes it satisfying for young idealist students to use for their project. Once the algo is perfected it can also be used for Face detection in kegger pictures.
Facebook has a big problem. The govt is asking them to police offensive images. They cannot hire enough humans to do it so the
Re: (Score:2)
CS departments and research IT are woefully underfunded. Partnering up with a big company is the new business model for funding research, whether it's nVidia, Microsoft or Facebook, they all are competing to get into modern research by giving away valuable resources (eg. cloud computing and physical hardware) to be able to get their hands on the datasets.
How are they going to monetize this? (Score:3)
They will not do ANYTHING out of the goodness of their (lack) of heart.
Re: (Score:2)
I've seen this movie ... (Score:2)
... and it didn't turn out well.
I worked for Mobil Oil.
They made so much money, they had a cash store (ca. 1986) that was obscene and the shareholders wanted them to do something with it that would make more money.
Mobil bought out an insurance company, went self-insured, and sold policies to any and all.
They also went into the land-grabbing business and built Reston, Va. from the ground up.
They bought Montgomery Ward, too.
They folded shortly after I retired from there.
--
When companies step away from their c
Isn't Facebook a Social Media Company? (Score:2)
FB and MRI scans? (Score:2)
Newfangled k-space sampling? (Score:1)
Every MR protocol is an engineering protocol that balances resolution (voxel size), signal-to-noise, field of view, and scanning time. If you increase resolution, you generally have to decrease your field of view, decrease SNR, or increase your scanning time. It's easy to make a 6 second scan - in fact, we do it every time we scan with what's called a localizer - a wide FOV, low resolution image that the techs use to orient the diagnostic images to be acquired. The localizer is not diagnostic, but I've caug
Not a new idea (Score:2)