REGISTER | LOGIN
Breaking News
Comments
Newest First | Oldest First | Threaded View
sixscrews
User Rank
Author
Re: Radiology?
sixscrews   5/9/2017 5:14:57 PM
NO RATINGS
Yah.

David Ashton
User Rank
Author
Re: Radiology?
David Ashton   5/9/2017 5:02:52 PM
NO RATINGS
@Sixscrews....from what you've said the problem is less about bad diagnosis and more about a system that wants to take as much as possible of your money and give as little as possible back?

sixscrews
User Rank
Author
Re: Radiology?
sixscrews   5/9/2017 4:06:43 PM
NO RATINGS
Point taken, Junko.

However, even I could tell there was something wrong with my left hip (besides the fact that I couldn't walk more than a half block) - the line corresponding to cartilage thickness was about 1/10th the size of the other hip.

This, perhaps, could have been flagged by a simple comparative algorithm and passed on to the radiologist before s/he looked at the image.  The difference in itself might not have been significant - we all have asymmetries in our bodies - but it would have merited a second look.

But I got a reading 'mild arthritis in left hip' - while an MRI three months later came up with 'severe arthritis in left him including bone marrow edema.'

Granted, MRI results often overstate artiritic changes in joints but these results do not compute.

ss/wb

junko.yoshida
User Rank
Author
Re: Radiology?
junko.yoshida   5/9/2017 3:25:30 PM
NO RATINGS
Yes, Infervision CEO also talked about "traditional computer-aided machine vision programs, such as R2." Those have been around for a couple of decades, he said, as the wiki entry you posted shows. The question now becomes how a new generation of deep learning-based screening would stack up against human doctors.

realjjj
User Rank
CEO
Re: Radiology?
realjjj   5/9/2017 3:06:02 PM
NO RATINGS
https://en.wikipedia.org/wiki/Computer-aided_diagnosis

junko.yoshida
User Rank
Author
Re: Radiology?
junko.yoshida   5/9/2017 2:58:28 PM
NO RATINGS
Hi! You might be onto something here. I can see how AI/Deep Learning would be useful in "filtering" information, but a lot of things doctors have to decide on -- in terms of diagnosis -- are judgment calls, I think.

How would a machine learn those borderline cases? Machine can be only as good as judgement calls doctors made previously, because everything machines learn is based on data they are fed with.

sixscrews
User Rank
Author
Radiology?
sixscrews   5/9/2017 2:21:23 PM
NO RATINGS
'Deep Learning' might be nice in Radiology.

I had an x-ray taken of my left hip in 2010 - the radiologist's report came up 'mild arthritis in left hip.'

I took a copy to an Orthopod MD I knew - outside my insurance network  so I paid in cash (this is the USA - Land of the Free) - and she said 'your left hip is trashed' (or something like that).  It took me six months to convince my medical insurance to get a left hip replacement.

Incredible results - improved my life 10^2.

So, what use an arbitrary human radiologist?  Some know their stuff, some are hacks in it for the money.

But the more apropos question is 'What use is a "deep learning" radiologist?'

wb



Like Us on Facebook
EE Life
Frankenstein's Fix, Teardowns, Sideshows, Design Contests, Reader Content & More
Martin Rowe

Test Tool Finds Ethernet Wiring Errors
Martin Rowe
Post a comment
When my house was renovated several years ago, I had the electrician install network outlets in numerous places, then run the LAN cables to a wiring closet. But he didn't document the ends ...

Martin Rowe

Local Electronics Store Supplies Engineers and Hobbyists
Martin Rowe
5 comments
Rochester, N.Y. — Tucked away in this western New York city known for its optics is Goldcrest Electronics, a local store that's supplied businesses and individuals with electronic ...

Martin Rowe

How to Transform a Technology University (Book Review)
Martin Rowe
1 Comment
The Presiding Genius of the Place by Alison Chisolm. WPI, Worcester, Mass., 234 pp., 2016. Engineers love to discuss, and often criticize, engineering education. They often claim ...

Max Maxfield

Aloha from EEWeb
Max Maxfield
Post a comment
Just a few minutes ago as I pen these words, I posted this blog about this month's Cartoon Punchline Competition over on EEWeb.com.