An Automatic Computer-Aided Detection Scheme for Pneumoconiosis on Digital Chest Radiographs

Abstract

This paper presents an automatic computer-aided detection scheme on digital chest radiographs to detect pneumoconiosis. Firstly, the lung fields are segmented from a digital chest X-ray image by using the active shape model method. Then, the lung fields are subdivided into six non-overlapping regions, according to Chinese diagnosis criteria of pneumoconiosis. The multi-scale difference filter bank is applied to the chest image to enhance the details of the small opacities, and the texture features are calculated from each region of the original and the processed images, respectively. After extracting the most relevant ones from the feature sets, support vector machine classifiers are utilized to separate the samples into the normal and the abnormal sets. Finally, the final classification is performed by the chest-based report-out and the classification probability values of six regions. Experiments are conducted on randomly selected images from our chest database. Both the training and the testing sets have 300 normal and 125 pneumoconiosis cases. In the training phase, training models and weighting factors for each region are derived. We evaluate the scheme using the full feature vectors or the selected feature vectors of the testing set. The results show that the classification performances are high. Compared with the previous methods, our fully automated scheme has a higher accuracy and a more convenient interaction. The scheme is very helpful to mass screening of pneumoconiosis in clinic.

Support Vector Machine Model for Diagnosing Pneumoconiosis Based on Wavelet Texture Features of Digital Chest Radiographs

“…To build and evaluate a more clinically meaningful model for the diagnosis of pneumoconiosis, DR samples of patients with stage III pneumoconiosis were excluded from our study. With only DRs of patients with stage I and II pneumoconiosis and of normal individuals, the SVM classifier built in this work got a rather high performance with an accuracy of up to 92.0 % and an AUC value of up to 0.97, which was not inferior to or even superior to those developed by using DR samples including DRs of stage III pneumoconiosis in other studies [4,6].…”

“…Some studies [2,18,19] detected the small rounded or somewhat irregular opacities on a chest radiograph and then classified the radiograph as a normal radiograph or a pneumoconiosis one according to the standardized system for classifying radiographic abnormalities of pneumoconiosis as established by the International Labor Organization. Other studies [3][4][5]20] took the texture analysis on the chest radiographs to diagnose pneumoconiosis. Texture features used in these studies included statistical texture features from the gray-level histogram and gray-level co-occurrence matrix, power spectrum, and frequency on the chest radiographs.…”

“…Studies investigating CAD for pneumoconiosis date back to the 1970s, with a recent revival of interest in the late 1990s [2,3]. Yu et al [4] detected pneumoconiosis using the gray-level histogram features and the co-occurrence matrices features on digital radiographs (DRs). Their support vector machine (SVM)-based models gave an overall accuracy of up to 89.2 % and an area under the receiver operating characteristic (ROC) curve of up to 0.978 when tested on 300 normal DRs and 125 pneumoconiosis DRs.…”

Development of CAD based on ANN analysis of power spectra for pneumoconiosis in chest radiographs: effect of three new enhancement methods

“…This is because the classification performance with the previous method was affected zone of the lung [22], or abnormal ROIs [28] included various subcategories, shapes, and sizes that it was easy for radiologists to classify as pneumoconiosis on chest radiographs. Therefore, for improved classification performance, typical texture patterns (each subcategory, shape, and size) were enhanced by texture features of the GLCOM, RLM.…”

Endobronchial fibroma in a pneumoconiosis patient with a history of tuberculosis: A case report and literature review

“…In 2008, a chest radiograph was performed as part of a health evaluation, which identified small rounded and irregular opacities in bilateral upper lobes. The patient was diagnosed with early-stage pneumoconiosis without presenting symptoms, based on the diagnostic criteria of pneumoconiosis (20,21). One year prior to admission, the patient had experienced a cough with sputum production.…”