Automated detection of infected area in lungs

Abstract

Interpretation of medical image is often difficult and time consuming, even for experienced physicians. Medical imaging plays an important role in detecting infected area in lungs. Significant efforts within the field of radiation oncology have recently been centered on the ability to automatically detect lung infection in a breathing patient in real-time during radiation treatment using digital x-ray technology and image processing. The motivation of such a goal is to improve radiation treatments, possibly leading to an increase in survival rates. Diagnosis of X-rays can be used as an initial step in detecting infected area in lungs. This paper describes a pulmonary region in X-ray images. We developed a automated system for the X-ray images. The computer scans and marks suspicious looking areas in the image. Radiologists can then focus on those areas and decide if further evaluation is needed. To improve the diagnosis accuracy of this system we introduced a pre-processing stage which involves adjustment of the intensity and conversion to gray scale of X-ray image. Gaussian filter is used to remove the false structure. Then we process the digitized image and extract the lungs from X-ray image. Then we apply blood fill algorithm for lung region, localization of suspected infected area. As a result infected area of lung is detected.