296
By Anum Memon, Syeda Munazza, Sheeraz Memon
Keywords: CAD, Chest Radiograph, JSRT dataset, Segmentation.
In this paper, an automatic segmentation method that detects lung boundaries and bypass the irrelevant information is presented. The automatic screening of chest radiographs is crucial for the rural areas where timely presence of expert radiologists is difficult. An automatic system can help save many lives. In digital image processing of chest radiographs, to detect abnormal indications, the system must be able to perform an accurate segmentation of lung area from an image because the major abnormal symptoms lies in the lung region. The designed system perform an automatic segmentation of lung region from chest radiographs through various stages such as image acquisition ,filtering , enhancement, morphological image processing, edge detection and segmentation. The system is tested on the Japanese Society of Radiological Technology (JSRT) and acquired deep information from the Chest Specialist and Radiologist for the better results.