Covid-19 chest CT-scan Dataset

Provide a large (1 terabytes) Covid-19 chest CT-scan Dataset

Dataset

Developed by
License
Other
A specific licence is required and is available on FIDAC website and in the Documents section below.

Main Characteristic

A total of 5944 patients were included from the 22 centers aggregated into 8 regions. Reason for the CT examination was provided for all the cases. RT-PCR results were provided in 5574 cases (93.77%) with a positive test in 44.6% of the cases.

The dataset is approximately 1 terabyte and 4 million images.

 

Research areas
Collaborative AI Explainable AI Verifiable AI
Technical Categories
Computer vision Machine learning
Business Categories
Healthcare
Last updated
10.12.2021 - 22:23
Detailed Description

The data collected were chest CT images in DICOM format associated with the following metadata:

  • patient age and sex,
  • originating facility identifier and originating facility region (if more than 100 cases, otherwise "other"),
  • time from symptom onset to CT scan: <3 days, 3-7 days, 8-21 days, and >21 days,
  • indication for CT scan ("low symptoms COVID-19 ", "suspicion of COVID-19 in a oxygen dependent patient ", "follow-up" and "other"),
  • final RT-PCR test result: positive, negative or not performed,
  • the normalized CT report status: “classic/probable COVID-19”, “intermediate for COVID-19”, “non-COVID” and “normal”

Trustworthy AI
The dataset has been irreversibly de-identified. Data reflect only the population tested to COVID-19 by CT-scans and does not contain any kind of ethnic information.
GDPR Requirements
The dataset has been irreversibly de-identified. It is GDPR-compliant.