Optimization of upper respiratory tract cancer diagnosis method based on exhaled breath gas analysis

Abstract

Listen

Translate article

Introduction. Head and neck tumors comprise about 7 % of all malignant neoplasms. In the head and neck area, tumors are usually located on the tongue (25–40 %) and floor of mouth (15–20 %). In the majority of cases, diagnosis, especially at early disease stages, is based on clinical and histopathological evaluation of tumor process. However, recently development and implementation of non-invasive techniques of early diagnosis of upper respiratory tract tumors through detection of pathognomonic volatile tumor markers in the exhaled air has become topical.Aim. To evaluate diagnostic accuracy of sensory gas analysis device and artificial neural network for examination of exhaled gas samples from patients with oropharyngeal, laryngeal, laryngopharyngeal cancer and to establish the optimal conditions for sampling.Materials and methods. The study included 28 patients with oropharyngeal, laryngeal, laryngopharyngeal cancers and 25 healthy volunteers. The proposed technique is based on analysis of samples of exhaled gas from the studied individuals using a diagnostic device developed by the authors. The device detects volatile compounds in the exhaled air using a set of semiconductor sensors with subsequent analysis by a neural network. The exhaled air was sampled using two methods: in the morning in the fasted state before daily hygienic procedures and physical activity (prepared samples) and in the context of everyday life, nutrition and hygiene without restrictions before sampling (non-prepared samples).Results. Based on the signals from the sensors, the neural network classified and detected patients with malignant tumors. Accuracy of the prepared samples from healthy volunteers and patients with oropharyngeal, laryngeal, laryngopharyngeal cancers was 79.17 %, of non-prepared – 84.09 %.Conclusion. High diagnostic accuracy of the developed technique of non-invasive diagnosis of malignant tumors of the oropharyngeal, laryngeal, laryngopharyngeal areas using exhaled air samples which does not require special preparation of the studied samples was demonstrated.