Zebra Medical Vision

Current Time Inside Cache Tag Helper: 11/23/2020 11:36:35 AM and Model.VendorId = 78130

Zebra Medical Vision's Current COVID-19 Response and Solutions

The following information was supplied by the vendor and has not been validated by KLAS.

Identify and track development of GGO lesions, offer insights into COVID-19 severity, enable quick and effective diagnosis, triage, and evaluation of patients

  • While laboratory testing has served as the first line of diagnostics during the COVID-19 crisis, these tests have limited accuracy and do not indicate the state of disease progression. CT scans have been proven to be the most accurate tool so far for screening COVID-19 (98% accuracy vs. 71% of the PCR test).
  • Zebra is working on software for a machine-learning algorithm that analyzes CT scans. That software will offer substantially higher detection rates while conducting a comparative analysis of multiple CT scan images of the same patient in order to measure the development of the virus.
  • We are leveraging a patented algorithm that Zebra developed in the past for detecting ground glass opacities (GGO), which are imaging features of COVID-19 pneumonia. The patented technology provides segmentation of predefined features in a 2D slice of a 3D anatomical image, including GGO.
  • Zebra researchers and engineers are currently using clinical data from thousands of medical images to train the AI model of the COVID-19 solution. The algorithm will be validated against confirmed COVID-19 CT scans from newly created databases around the world.
  • By identifying and tracking the development of these GGO lesions, the software will offer key insights into disease severity and enable doctors to diagnose, triage and evaluate patients swiftly and effectively.
  • We do not believe this will replace the PCR test, but it may have substantial value since deciding which suspected COVID-19 patients should be hospitalized is challenging due to the very limited resources that the public system has. The vast majority of COVID-19 patients are mild, so having this automated tool will help make critical decisions on whether or not to hospitalize based on a COVID-19 risk score driven by the analysis of the CT images.
  • The proof of concept will be available for assessment by mid-April.

Current Time Inside Cache Tag Helper: 11/23/2020 11:36:35 AM and Model.VendorId = 78130