Use of algorithms for semi-automatic quantification of pulmonary fibrosis and emphysema

Abstract

Obtaining a semi-automatic quantification of pathologies found in the lung, through images of high resolution computed tomography (HRCT), is of great importance to aid in medical diagnosis. Paraccocidioidomycosis (PCM) is a systemic disease that affects the lung and even after effective treatment leaves sequels such as pulmonary fibrosis and emphysema. It is very important to the area of tropical diseases that the lung injury be quantified more accurately. In this stud, we propose the development of algorithms in computational environment Matlab® able to objectively quantify lung diseases such as fibrosis and emphysema. The program consists in selecting the region of interest (ROI), and through the use of density masks and filters, obtaining the lesion area quantification in relation to the healthy area of the lung. The proposed method was tested on 15 exams of HRCT of patients with confirmed PCM. To prove the validity and effectiveness of the method, we used a virtual phantom, also developed in this research. © 2013 Springer-Verlag.