A computer-aided diagnosis approach for emphysema recognition in chest radiography

Abstract

a b s t r a c tThe purpose of this work is twofold: (i) to develop a CAD system for the assessment of emphysema by digital chest radiography and (ii) to test it against CT imaging. The system is based on the analysis of the shape of lung silhouette as imaged in standard chest examination. Postero-anterior and lateral views are processed to extract the contours of the lung fields automatically. Subsequently, the shape of lung silhouettes is described by polyline approximation and the computed feature-set processed by a neural network to estimate the probability of emphysema.Images of radiographic studies from 225 patients were collected and properly annotated to build an experimental dataset named EMPH. Each patient had undergone a standard two-views chest radiography and CT for diagnostic purposes. In addition, the images (247) from JSRT dataset were used to evaluate lung segmentation in postero-anterior view.System performances were assessed by: (i) analyzing the quality of the automatic segmentation of the lung silhouette against manual tracing and (ii) measuring the capabilities of emphysema recognition. As to step i, on JSRT dataset, we obtained overlap percentage (˝) 92.7 ± 3.3%, Dice Similarity Coefficient (DSC) 95.5 ± 3.7% and average contour distance (ACD) 1.73 ± 0.87 mm. On EMPH dataset we had ˝ = 93.1 ± 2.9%, DSC = 96.1 ± 3.5% and ACD = 1.62 ± 0.92 mm, for the postero-anterior view, while we had ˝ = 94.5 ± 4.6%, DSC = 91.0 ± 6.3% and ACD = 2.22 ± 0.86 mm, for the lateral view. As to step ii, accuracy of emphysema recognition was 95.4%, with sensitivity and specificity 94.5% and 96.1% respectively. According to experimental results our system allows reliable and inexpensive recognition of emphysema on digital chest radiography.

Lung field segmentation using weighted sparse shape composition with robust initialization

“…We compared our method with many other methods or systems. 11,14,17,[26][27][28]37 The performances of all methods are summarized in Table V. Considering that the metrics used in different methods or systems were different in literatures, we only listed the most common metrics (overlap, ACD, and DSC) in Table V.…”

A review on lung boundary detection in chest X-rays

“…The edges are then completed using curve fitting techniques. A recent effort [45] was aimed at the automated computation of emphysema utilizing the shape of lung fields in both frontal and lateral chest radiographs. The lung boundary is modeled as a closed fuzzy curve and estimated by self-organizing networks [58] (Fig.…”

“…In [45, 72, 73], researchers utilized geometrical features extracted from lung boundaries to automatically detect emphysema. The other structural abnormality is cardiomegaly which is a medical condition caused by high blood pressure or coronary artery disease.…”

Chest Radiography: New Technological Developments and Their Applications

“…Therefore, CAD approaches for texture analysis could benefit from inclusion of digitally bone suppressed images or dual energy acquired images. 75,76 Other research focuses on the computerized detection of emphysema in chest radiographs, 77,78 detection of pneumothorax, 79 and the automatic detection of catheters and tubes. [80][81][82] Also CAD software for the detection of vertebral compression fractures on lateral examinations has been developed.…”