Diagnosis of Systolic Heart Failure Disease with an Electronic Nose

Electronic nose technology is attracting attention with its diagnostic applications in the healthcare field. In this study, respiratory samples of individuals with systolic heart failure (HFrEF) were analyzed using an electronic nose device to investigate the diagnostic feasibility for this disease. A total of 275 breath samples were collected from 29 patients and 31 healthy volunteers followed in a cardiology clinic. Classification using support vector machines (SVM) yielded an average accuracy rate of 85.21%. The simplicity of the statistical features used in the classification, combined with the low computational complexity, increases the method’s practicality. This study demonstrates that, unlike existing imaging and laboratory techniques, electronic nose technology can be considered a non-invasive, rapid, and cost-effective alternative for diagnosing heart failure, particularly notable for its potential to contribute to early diagnosis.