Code related to the article "Automatic Orientation Identification of Pediatric Chest X-rays".
This article presents an automatic method to identify four orientations in pediatric radiographic images of the chest. The identification and correction of CXR plays an important role in PACS, CAD, and image registration systems, and can influence its performance. Our method uses the analysis of structural characteristics and statistics of CXR pixel intensity patterns in a resource extraction and selection scheme and then a decision tree machine learning classifier. We used three different databases that evaluated using both intra-database and inter-database tests. Our results showed high assertiveness, with our method being able to achieve accuracy of at least 99.4%, while 100% was achieved on the test set on one of the databases.
- Guangzhou : https://www.kaggle.com/paultimothymooney/chest-xray-pneumonia/version/2
- JSRT : http://db.jsrt.or.jp/eng.php
The main requirements are listed below:
- Matlab version 2014a or later
- Statistics and Machine Learning Toolbox
- Imaging processing and computer vision
This browser does not support PDFs. Please download the PDF to view it: Download PDF.