client logo
Version: 1.0.0 | Published: 8 Oct 2024 | Updated: 229 days ago

COVID-19 Detection from Chest X-Rays using Deep Learning

Dataset

Documentation

Description:
COVID-19 is a pandemic having devastating implications on healthcare systems globally. Evidence shows that COVID-19 infected patients with pneumonia may present on chest x-rays with a pattern that is difficult to characterise using only the human eye. Therefore, artificial intelligence (AI) techniques using deep learning, which can consistently identify infected patients from non-infected ones given a radiographic examination of the patient, can be used as a reliable diagnostic tool. Considering chest x-rays are one of the most commonly performed radiological studies (coupled with the near universal availability of testing machines), applying AI techniques on them could prove to be valuable for COVID-19 diagnosis during clinical management. We therefore aim to establish a reliable diagnostic tool based on a deep-learning framework for the screening of patients who present with COVID-19 related abnormalities on chest x-rays. Over the course of 7 months we will build a dataset using open source data which are freely available, as well as with de-identified patient data collected from health institutions in Pakistan. Using this dataset, a deep learning model will be trained, which would be able to accurately screen patients who present with abnormalities relevant to COVID-19 in their radiographic examination. This tool will ultimately aid in expediting the diagnosis and referral of COVID-19 patients, resulting in improved clinical outcomes. For further information, see: https://www.ed.ac.uk/usher/respire/covid-19/covid-19-detection-chest-x-rays

Coverage

Spatial:
Pakistan
Typical Age Range:
0-150

Provenance

Temporal

Accrual Periodicity:
Static
Start Date:
01 August 2020
End Date:
31 January 2021
Time Lag:
Not applicable

Accessibility

Access

Access Service:
Access is managed on a project-by-project basis. Contact the RESPIRE team.
Delivery Lead Time:
Not applicable
Jurisdictions:
PK
Data Controller:
RESPIRE
Data Processor:
RESPIRE

Usage

Resource Creators:
RESPIRE Collaboration

Format and Standards

Vocabulary Encoding Schemes:
LOCAL
Languages:
en
Formats:
text

Observations

Statistical Population
Population Description
Population Size
Measured Property
Observation Date
Findings
1
Count
31 January 2021