Detection of tuberculosis using improved gradient driven textures and secured qr pattern generation for chest xray images

Abstract

The recent resurgence of lung related diseases has dramatically influenced the biomedical field. Analyzing the lung region continues to be a challenge for medical professionals due to its inherent statistical characteristics. The accuracy measures inadequacies and computational complexity inherent in existing CAD systems have driven the development of new biomedical image processing model which can support wide range of image data sets. In addition to this, with the invention of cloud assisted biomedical system and associated digitalized data transmission - the data security is becoming most important than ever and also difficult task to accomplish. Among other diagnostic measurement the detection of tuberculosis from input lung images has become progressively more important and other lung image analyzes also emerging steadily in recent years. In general, computed Tomography (CT) and chest X-ray (CXR) images play a significant role in tuberculosis screening and explore different anatomical lung region changes within image. Even though there are various systems available for diagnosing lung abnormalities, computer-aided diagnosis (CAD)-based identification of lung diseases and automated tuberculosis diagnosis in the subsequent stages face specific problems due to homogeneities, anatomical shape variations, and uneven lung boundary conditions that arise in X-ray imaging. newline