Detection of abnormality in chest X-rays using deep learning

Not Applicable

Recruiting

• Conditions: CXRs images with and without definite diagnosis.; CXR; Deep learning; Detection; Abnormality

• Registration Number: TCTR20200925003
• Lead Sponsor: Faculty of Medicine, Khon Kaen University

Brief Summary

Not available

Detailed Description

Not available

Study Timeline

• Study Start: 2020-09-25
• Study Completion: 2021-12-31
• Last Update Posted: 19 August 2024

Recruitment & Eligibility

• Status: Recruiting
• Sex: All
• Target Recruitment: 1000

Inclusion Criteria

CXR with pulmonary and cardiac diseases.
CXR without abnormality.

Exclusion Criteria

No image in local PACS

Study & Design

• Study Type: Observational
• Study Design: Not specified

Primary Outcome Measures

Name	Time	Method
Detection of abniormal CXR At the end of the project Accuracy/ ROC curve

Secondary Outcome Measures

Name	Time	Method
ocalization of abnormality At the end of the project Accuracy/ Dice index