Two-stage pneumonia detection through object detection and patch classification

Two-stage pneumonia detection through object detection and patch classification
Title:
Two-stage pneumonia detection through object detection and patch classification
Other Titles:
IEEE-EMBS International Conference on Biomedical and Health Informatics (BHI)
DOI:
Publication URL:
Keywords:
Publication Date:
01 May 2019
Citation:
Abstract:
Pneumonia is a common disease that affects mil-lions of people globally, typically requiring the expertise of experienced radiologists to diagnose the disease. Previous methods learn to detect pneumonia through comparison of global spatial features of other diseases. Our proposed pneumonia detection method uses the identification of regions of interest within chestX-ray images and then performs classification of the regions.Experiment results show that our two-stage object detection and classification can achieve higher recall and accuracy.
License type:
PublisherCopyrights
Funding Info:
Description:
(C) 2019 IEEE
ISBN:

Files uploaded:
File Size Format Action
There are no attached files.