NYU Langone Health and Meta have developed a new type of MRI that dramatically reduces the time it takes to complete a scan using artificial intelligence. CBS News correspondent Anne-Marie Green reports. CBS News Streaming Network is the premier 24/7 news streaming service from CBS News and Stations, available for free to anyone with Internet access. The CBS News Streaming Network is your destination for breaking news, live events and original reporting locally, nationally and around the world. Launched in November 2014 as CBSN, the CBS news streaming network is available live in 91 countries and on 30 digital platforms and apps, as well as on CBSNews.com and Paramount+. Subscribe to CBS News YouTube Channel: Watch CBS News: Download the CBS News App: Follow CBS News on Instagram: Like CBS News on Facebook: Follow CBS News…
So great! I hope they get it in Australia. My scans are 1 hr ( Have one tomorrow) for a tumor, fibrous dysplasia left femur (unicameral) with contrast 🤦♀️
AI gonna take you're job!!!
That’s been around for years; GE pioneered that with AIrecon.
The billion dollar question is why hasn’t this doctor been replaced by AI too.
So then, costs twice as much?
Glad to see AI is becoming useful.
I was thinking of this to speed up the scans. This should include other scans too. I'm waiting 4 months for a CT scan.
They mention that it's 'faster' and has 'clearer images' but both of those are very much runners up to the most important quality, ACCURACY. Let's be clear, this is functioning by filling in gaps in the data so it looks better, but is that compromising accuracy? Are potentially vital health indications getting scrubbed out as noise? Seems odd that the question of accuracy is never addressed.
WOW, this is good news!
Scary.
Google DeepMind, IBM Watson some of Ai done this for years
I can see the difference, but clinically, what does it mean? Now you have a computer doing more processing of an image. Can it help in making a better diagnosis or can it increase the risk of misdiagnosis because the AI is enhancing an image in a way that does not correspond to the real physiological state of the patient? I think the diagnosis from both images would be the same.
As an analogy, consider that wearing yellow lens glasses on a cloudy day helps most people see better. But wearing coloured lenses also distorts our colour perception — even on a black and white image — as someone who used to used black and white film will tell you. I wish the report explained how these improvements are being made. For example, is it using an interpolation algorithm?
Only had an MRI once, well twice actually when the first images were somehow corrupted. And it lasted roughly 20 minutes, but was for my shoulder so my head was right in the machine without any sticking out and yeah there's a level of anxiety there I didn't realize I had. I quickly found out that closing my eyes so I didn't realize how tight the area is did help a lot, and like the woman says just day dream and take your mind somewhere else. Having something to speed up this process would be fantastic though, just so long as it's not creating false data by extrapolating what it believes things should look like.
So happy to see this..
Meta is not doing this for anything good… I’m going to be completely candid that should be highly concerning
Wonder how long it will take for health insurance companies to decide they won’t cover it
Now if only ai could fix my acl
yeah we are just now realising how useless doctors are and over payed by a lot
this is my best hope for AI technology. Some how we just need guardrails on the dark side.
So the algorithm infers pixels rather than raw data. precise enough i trust