Support vector machine-based diagnosis of Tuberculosis

Tuberculosis is one of the common death-causing diseases in the world and possess a threat to the health of the world population. This study aims to elaborate on the feasibility of developing a setup of a combination of hardware and software, using respiratory sounds for diagnosis of Tuberculosis. This work adds to the existing literature body and for the facilitation of the scientific community. This proposed method has efficiently achieved the diagnosis of tuberculosis using Lung sound processing. It can be used to facilitate the treatment of patients which will help physicians and healthcare organizations. The designed system achieved the highest accuracy of 100%, a specificity of 100%, and a sensitivity of 100% using only four features and an SVM classifier 10-fold cross-validation.