🔍
Task-aware and Anatomy-specific Cone-beam Computed Tomography
Case ID:
C15787
Report of Invention:
4/4/2019
Web Published:
10/18/2019
Unmet Need
During minimally invasive orthopedic surgery, metallic instruments such as metallic wires and screws are implanted to restore bone morphology and restore function. Currently, the clinically acceptable placement of these instruments can only be verified reliably on post-operative CT images. Because post-operative verification prohibits immediate repositioning of the implant, there is a need for intra-operative 3D cone-beam CT (CBCT) that will allow immediate intra-operative verification and consequently immediate repositioning, thus removing the need for revision surgery. While the robotic CBCT-enabled C-arm X-ray systems are already installed in many state-of-the-art operating suites, they are not currently used for intra-operative verification due to the severe image artifacts in tomographic reconstructions that inhibit meaningful interpretation. These artifacts arise from discrepancies between the physical effects that govern image formation and the assumptions that must be made to enable algorithmic 3D reconstruction. The extent of mismatch between imaging physics and algorithmic assumptions depend on the 3D scene that is imaged and the view onto the scene. As a consequence, selecting views that do not violate the imposed assumptions as heavily will yield improvements in reconstructed image quality.
Technology Overview
Based on the above observations, the inventors have developed an improvement on CBCT that utilizes the mobile C-arm for intra-operative 3D CBCT while also providing accurate and clinically viable images. Specifically, they have built upon traditional short-scan CBCT trajectory by extending the planer trajectory and autonomously adjusting out-of-plane angulation. This allows for the C-arm source trajectories to be scene-specific and avoid unusable images characterized by beam hardening, photon starvation, and noise. This algorithm is performed in real-time using a deep convolutional neural network that regresses a detectability-rank derived from imaging physics. Not only does this method provide clinically viable reconstructed images in a task-aware and patient-specific manner, it also is performed in real-time utilizing the mobile C-arm x-ray system.
Stage of Development
The inventors have developed an in silico test bench for task-aware CBCT imaging protocols and have retrospectively evaluated their algorithm on semi-anthropomorphic phantoms. They plan on testing the system for accuracy and complexity assessments.
Patent Information:
Title
App Type
Country
Serial No.
Patent No.
File Date
Issued Date
Expire Date
Patent Status
A MACHINE LEARNING MODEL TO ADJUST C-ARM CONE-BEAM COMPUTED TOMOGRAPHY DEVICE TRAJECTORIES
PCT: Patent Cooperation Treaty
United States
17/634,597
11,992,348
2/11/2022
5/28/2024
5/24/2041
Granted
A MACHINE LEARNING MODEL TO ADJUST C-ARM CONE-BEAM COMPUTED TOMOGRAPHY DEVICE TRAJECTORIES
CON: Continuation
United States
18/659,587
5/9/2024
Pending
Direct Link:
https://jhu.technologypublisher.com/technology/36711
Inventors:
Category(s):
Technology Classifications > Computers, Electronics & Software > Image Processing & Analysis, Technology Classifications > Computers, Electronics & Software > Algorithms, Technology Classifications > Computers, Electronics & Software > Robotics,
Get custom alerts for techs in these categories/from these inventors:
Subscribe for JHTV Updates
For Information, Contact:
Lisa Schwier
lschwie2@jhu.edu
410-614-0300
Save This Technology:
Bookmark this page
Download as PDF
JHTV Home
|
Search
|
Login/Subscribe
2017 - 2022 © Johns Hopkins Technology Ventures. All Rights Reserved. Powered by
Inteum