Computerized Analysis of Pneumoconiosis in Digital Chest Radiography: Effect of Artificial Neural Network Trained with Power Spectra

Abstract

It is difficult for radiologists to classify pneumoconiosis with small nodules on chest radiographs. Therefore, we have developed a computer-aided diagnosis (CAD) system based on the rule-based plus artificial neural network (ANN) method for distinction between normal and abnormal regions of interest (ROIs) selected from chest radiographs with and without pneumoconiosis. The image database consists of 11 normal and 12 abnormal chest radiographs. These abnormal cases included five silicoses, four asbestoses, and three other pneumoconioses. ROIs (matrix size, 32 × 32) were selected from normal and abnormal lungs. We obtained power spectra (PS) by Fourier transform for the frequency analysis. A rule-based method using PS values at 0.179 and 0.357 cycles per millimeter, corresponding to the spatial frequencies of nodular patterns, were employed for identification of obviously normal or obviously abnormal ROIs. Then, ANN was applied for classification of the remaining normal and abnormal ROIs, which were not classified as obviously abnormal or normal by the rule-based method. The classification performance was evaluated by the area under the receiver operating characteristic curve (Az value). The Az value was 0.972 ± 0.012 for the rule-based plus ANN method, which was larger than that of 0.961 ± 0.016 for the ANN method alone (P ≤ 0.15) and that of 0.873 for the rule-based method alone. We have developed a rule-based plus pattern recognition technique based on the ANN for classification of pneumoconiosis on chest radiography. Our CAD system based on PS would be useful to assist radiologists in the classification of pneumoconiosis.

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.…”

“…Therefore, because we obtained more information on abnormal and normal lungs, we developed a CAD system for the distinction between normal and abnormal patterns in pneumoconiosis using the ANN trained with the power spectrum (PS) values [28]. …”

The Reliability of Artificial Neural Network in Locating Minor Apical Foramen: A Cadaver Study

“…Also it modifies its behavior (trains) by adjusting the strength or weights of the connections until its own output converges to the known correct output. Once trained, the network can evaluate a new case of input values by applying the weights learned from the data set with which it was trained (45). The network can generalize from previous cases to evaluate cases not previously seen.…”

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

“…Using a similar approach, Xu et al [5] distinguished 175 pneumoconiosis DRs from 252 normal DRs, giving a 95.5 % overall accuracy. In another investigation [6], power spectra of chest DRs which implied image texture were obtained through Fourier analysis and were input into an artificial neural network with the backpropagation algorithm to detect pneumoconiosis, resulting in an area under ROC curve (AUC) of 0.961. In our previous study [7], the use of texture features extracted through graylevel histogram and co-occurrence matrix of chest DRs and artificial neural network with the back-propagation algorithm also suggested a relatively high performance for the diagnosis of pneumoconiosis.…”