1- , mp.hosseini@srbiau.ac.ir
Abstract: (11368 Views)
Background: Chronic Obstructive Pulmonary Disease (COPD) is one of the most prevalent pulmonary diseases. Use of an automatic system for the detection and diagnosis of the disease will be beneficial to the patients' treatment decision-making process. In this paper, we propose a new approach for the Computer Aided Diagnosis (CAD) of the disease and determination of its severity axial CT scan images.
Methods: In this study, 24 lung CT scans in full inspiratory and expiratory states were performed. Variations in
the normalized pattern of the lungs' external parenchyma were exploited as a feature for COPD diagnosis.Subsequently, a Bayesian classifier was used to classify variations into two normal and abnormal patterns for the discrimination of patients and healthy individuals. Finally, the accuracy of the classification was assessed
statistically.
Results: With the proposed method, the lungs parenchymal elasticity and air-trapping
were determined quantitatively. The more this feature tended to zero, the more severe air-trapping and obstructive pulmonary disease is. By analyzing CT
images in the healthy and patient groups, we calculated the hard threshold for the diagnosis of the disease. Clinical results tested by the
mentioned method, suggested the effectiveness of this approach.
Conclusion: In regard to the challenges of COPD diagnosis, we propose a new computer-aided design which may be helpful to physicians for a more accurate diagnosis of the disease. Moreover, this severity scoring algorithm may be
useful for targeted disease management and risk-adjustment.