Hey Veeky Forums,
I am a physics major and have been looking into medical imaging and find the physics behind it fascinating. One thing that really captures my attention about it is the problem of automating the diagnosis from a medical image itself.
What is the current state of this field? What are some resources i should look into to help me solve this problem?
Other urls found in this thread:
www-03.ibm.com
cnn.com
twitter.com
Deep learning microscopy would be a good place to start. Apply deep learning techniques from there to medical imaging to increase chances of correct object recognition.
Signal processing and image processing pretty much cover it
I worked in a medical imaging lab at Vanderbilt a few years ago, and while they were developing computational tools to process medical images, there was not a lot of work being done on automated diagnosis.
Automated diagnosis with less and less human involvement will require much more computational power and resources, not just better programs.
IBM had a big puff piece about a week ago.
www-03.ibm.com
Watson can do more than play "Jeopardy"
>Deep learning microscopy
not OP, but what's that?
Diagnosing tissue samples (i.e. cancer) using AI.
Are there any other infographs like in OP's?
interested in this as well
bump
my lab in THz electronics is working on medical imaging applications
>THz electronics
i thought this turned out to be a meme
is it a meme user?
Checked quints
personally don't think so, but developments in the field have been slower than I think most people wanted.
what makes you think it's not a meme?
also, why have developments been slower than expected? where did most people think the field should be by now?
bump
Bump
currently work in a lab relying heavily on MRI/MRS imaing processing (for fat ). Wonderful tool perhaps the greatest diagnostical tool would currently available (imaging wise ). i can secong that typically we developt very specialized tool but i thinking it would be a monstrous task to develop deep learning.
maybe radio-oncology and endocrinology would be the prime field to use a algorithm
How did you get involved with that lab? What's your educational background
checked
