A Common Framework for the Extraction of ILD Patterns from CT Image

Abstract

Majority of the pulmonary diseases and their identification rely on geometric progression of lung spaces. Most common types of lung diseases include abnormalities categorised as Interstitial lung diseases (ILD) like sarcoidosis, idiopathic pulmonary fibrosis (IPF), malignant nodules, extrinsic allergic alveolitis (EAA) and honey comb structures from the infectious disorders is a very difficult task for diagnosis. For clinical practices, images are accumulated and stored in digital representation like MRI and CT to facilitate corresponding diagnosis. Some of the physicians can’t provide inadequacy in image parts which are known as (ROI) region of interest. Researchers converse at focusing on ROI coding to guarantee the use of multiple and randomly shaped ROI’s in image depicting the importance of ROI confining the background regions that can be exhibited by varying the levels of quality. This paper highlights the medical image ROI segmentation that delineates the diseased part using morphological algorithm. This paper addresses working on reliable methods for diagnosis and prognosis of the pulmonary diseases. Segmentation of ROI for the detection of CT lung pattern abnormalities likely nodules, sarcoidosis, IPF and honeycomb are done based on morphology in this research work. The techniques used to decoct medical information helps the radiologists for early diagnosis of ILD to figure out appropriate treatment.