Identifying honeycombing structure in HRCT lung images by high intensity pixel pattern

Lung diseases in most often results in fatality. Smoking, infections, and genetics are some of the factors that cause lung diseases. In 2014 alone 72489 Indian people were affected by lung disease among them 65672 people died. CAD system is of best help to doctors for diagnosis of lung diseases in higher degree of accuracy and timely treatment. High-resolution Computed Tomography (HRCT) is the popular and efficient method for diagnosing a number of lung diseases. Most of the diseases involving fibrosis of are diagnosed by the presence of Honey comb structures in lungs. The lungs affected with certain end-stage interstitial lung diseases (ILDs) are seen with fibrotic cystic changes and known as Honeycomb lung. Lung images of patients suffer from idiopathic pulmonary fibrosis (over 70%) and other causes of usual interstitial pneumonia (UIP), cystic bronchiectasis, pneumoconiosis and tuberous sclerosis show honeycomb structure. Image processing techniques play a significant role in the diagnosis of lung diseases. This paper presents a novel image processing method to identify the honeycomb structure in lung CT images using the pattern identification of the presence of the number of high intensity pixels with pixel intensity level >200 in the identified ROIs (Region of interest). The results obtained show the feasibility of the proposed method to diagnose the honeycomb structure of lung CT image.