A research team has proposed non-contrast thoracic chest CT scans as an effective tool for detecting, quantifying, and tracking COVID-19. As of March 16, the COVID-19 pandemic had a confirmed infection total of more than 170,000 people around the globe. The speed of transmission of COVID-19 has surprised the world and had a massive impact on people’s daily lives and the global economy.
To accelerate COVID-19 detection and support efforts to combat the epidemic, researchers from RADLogics, Tel-Aviv University, New York Mount Sinai Hospital and University of Maryland School of Medicine developed an AI-based approach designed to help identify infected patients and quantify disease burden by analyzing thoracic CT (Computer Tomography, aka CAT) exams.
Data sources for new epidemic diseases such as COVID-19 remain limited, as does expertise. Without knowing whether there are enough cases to achieve clinically meaningful learning in the early data collection stage, researchers are trying to modify and adapt existing AI models with initial clinical understanding to rapidly develop AI-based tools to deal with the COVID-19 challenge. The proposed system combines 2D and 3D analysis deep learning models and clinical understanding and was trained on data from available international datasets, including from infected areas in China. Researchers also used basic data augmentation techniques.

The team conducted several retrospective experiments on a testing set of CT scans from 157 patients from China and the US. The results show the developed deep learning image analysis system was able to develop classification points and detect suspicious COVID-19 thoracic CT features. Classification of coronavirus vs non-coronavirus cases was excellent: 0.996 AUC per thoracic CT studies (95%CI: 0.989-1.00) on Chinese control and infected patients with possible working points of 98.2% sensitivity and 92.2% specificity.
The system also uses 3D volume analysis to assess disease state and generate a “Corona score” metric that can be used to monitor disease progression or regression in patients.

Automated CT image analysis with deep learning algorithms has been shown to be effective in detecting, quantifying, and tracking coronavirus. The researchers propose that such highly accurate systems could be used to reliably exclude coronavirus-negative CTs in order to decrease the volume of scans sent to radiologists with confidence that positive cases are not being overlooked. This could enable more patient screening and earlier detection of positive cases to help with containment efforts.
Researchers say the study is currently being extended to larger populations.
The paper Rapid AI Development Cycle for the Coronavirus (COVID-19) Pandemic: Initial Results for Automated Detection & Patient Monitoring using Deep Learning CT Image Analysis is on arXiv.
Author: Herin Zhao | Editor: Michael Sarazen
Can you please help us to find a ct scan data set for covid 19.
We just released a small set of 100 segmented CT slices of COVID-19 at http://medicalsegmentation.com/covid19/
Can i get the link for your github handle
Thanks.
Thank you for an informative news, It’s very useful. Further more details about IBM Artificial Intelligence Helping Fight against COVID-19 on this video https://www.youtube.com/watch?v=iVPoWymbN6o&t=3s
Pingback: Die 5 wichtigsten Use Cases für Computer Vision | STATWORX
Pingback: Computer Vision | How does Computer Vision work
Pingback: Die 5 wichtigsten Use Cases für Computer Vision – Statworx
Pingback: The 5 Most Important Use Cases for Computer Vision – Statworx
Pingback: Die 5 wichtigsten Use Cases für Computer Vision - Statworx