CHANNELS

Aerospace

Automotive

Automated & Autonomous Vehicles

Battery & Electrification Technology

Defense

Drone TV

Electronics

Manufacturing & Prototyping

Materials

Medical

Motion Control

RF & Microwave Technology

Robotics & Automation

Sensors & IoT

Test & Measurement

Aerospace
Automotive
Automated & Autonomous Vehicles
Battery & Electrification Technology
Defense
Drone TV
Electronics
Manufacturing & Prototyping
Materials
Medical
Motion Control
RF & Microwave Technology
Robotics & Automation
Sensors & IoT
Test & Measurement
Aerospace & Defense
Search Results

'DeepDOF' AI Microscope Could Confirm Tumor Removal in Minutes

When surgeons remove cancer, one of the first questions is if all the cancerous tissue was removed. Researchers from Rice University  and the University of Texas have developed a new microscope that can quickly and inexpensively image large tissue sections, even during surgery, to find the answer. The microscope can rapidly image relatively thick pieces of tissue with cellular resolution, and could allow surgeons to inspect the margins of tumors within minutes of their removal. The deep learning extended depth-of-field microscope (DeepDOF) utilizes the AI technique of deep learning to train a computer algorithm to optimize both image collection and image post-processing.

Topics:
Imaging Medical Software

More From SAE Media Group

    Transcript

    00:00:00 when a patient goes through surgery the main goal of the surgery is to remove every like all the cancer cells in the patient but in order to do that you have to basically look at the tumor under a microscope so traditionally in order to do that you have to make these slides so they're actually pretty time consuming and costly to make so our

    00:00:23 project seeks to basically image the tissue directly so the faster and cheaper we can do this the easier it is for surgeons to complete the surgery the slide on the right it really takes a lot of expertise it takes a very expensive equipment and it also takes time especially during the at the time of the

    00:00:44 surgery you do not have the luxury to look at to get a lot of these slices traditionally the imaging equipment for example a camera or microscope is designed separately from any imaging processing algorithm that comes afterwards but deep dof is the first microscope that's

    00:01:02 designed with the post-processing algorithm in mind so it has this special optical element the face mask and a post-processing algorithm a unit and both of them are jointly trained using deep learning it helps the clinicians surgeons and pathologists to look at the irregular surfaces of the tissue samples

    00:01:24 directly without refocusing all the time and the way it works is that we have developed this deep learning architecture that can optimize both the optical design and image processing so that it can bring what is out of focus back into focus digitally

    00:01:44 we have demonstrated that this end-to-end jointly trained optics plus post-processing algorithm really works and it gives us better performance compared to when you have the separate components trained individually another component which is from the clinical perspective is that you can actually use this

    00:02:03 data-driven approach to build something very simple yet very effective and that really means we can bring these technologies into the more resource-constrained areas that cannot afford some of the more expensive equipment or personnel now they can also have access to these state-of-the-art imaging

    00:02:31 techniques