MSMikhail Saveliev

Mikhail Saveliev

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Mikhail Saveliev has a total of 17 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    The Potential of Using Multisensor Analysis of Volatile Organic Compounds in Exhaled Breath to Distinguish the Degree of Fibrosis in Patients with Diffuse Liver Disease (Pilot Study)

    by M. V. Kruchinina, M. F. Osipenko, R. A. Bukarev, I. E. Jahatspanian, M. A. Grevtsev, M. Yu. Savelyev, М. В. Кручинина, М. Ф. Осипенко, Р. А. Букарев, И. Э. Джагацпанян, et al.

    The aim of the study was to determine the potential of using volatile organic compounds in exhaled breath to differentiate patients with diffuse liver disease of various origins from healthy individuals, as well as from those with mild and severe liver fibrosis. Materials and Methods. A total of 87 patients with diffuse liver disease (DLD) of various etiologies were examined (mean age 51.7 ± 13 years, 45 men (51.7%) and 42 women (48.3%), and 36 individuals in the comparison group (conditionally healthy). The degree of liver fibrosis was determined using transient elastography on a FibroScan® 502 device (Echosens, France). Most patients with DLD had grade 0-II fibrosis (n=62, 71.3%) (group 1 with mild fibrosis), while grades III and IV fibrosis were established in 25 cases (group 2 with severe fibrosis). Liver cirrhosis was verified in 17 patients, mainly of mixed origin (metabolic + HCV + alcoholic) and Child-Pugh class A and B. All subjects underwent Multisensory analysis of volatile organic compounds (VOCs) in exhaled air using the domestically produced ARAMOS-7 Analyzer (“electronic nose”) (Stock Company Scientific Research and Production Association “Pribor”, St. Petersburg). A dataset of 141 information indicators from the examined subjects was mathematically processed using principal component analysis. The following classification methods were used to classify the groups: k-nearest neighbors (kNN), LR logistic regression, RF random forest, SVM support vector machine (SVM), and discriminant analysis (PLSDA) using latent structure regression. A validation procedure was conducted to verify the reliability of the classification. ROC analysis provided data on the diagnostic accuracy of the generated patterns for distinguishing between the groups. Results. Multisensory analysis of 141 information indicators demonstrated the ability to distinguish patients with diffuse liver pathology from individuals in the comparison group. Validation, conducted using a multiple iterative method of dividing the total sample into training and test sets, demonstrated diagnostic accuracy. 90%, sensitivity 89.4%, specificity 91.3%, AUC 0.93. The use of the estimated cross-validation method for exhaled breath VOC patterns yielded a diagnostic accuracy of 70.3%, sensitivity 76.2%, specificity 62.5%, and AUC 0.72 in differentiating between severe (F3-4) and mild (F1-2) fibrosis, indicating the acceptable quality of the developed model. Conclusion. The obtained results demonstrate the potential for continued research on VOCs using electronic nose technology, with an increased number of information indicators for the diagnosis and monitoring of patients with diffuse liver disease.

  • Determination of Halitosis by Exhaled Breath Analysis Using Semiconductor Metal Oxide Sensors and Chemometric Methods

    by Mikhail Saveliev, Andrey Volchek, Galina Lavrenova, Ol'GA Malay, Mikhail Grevtsev, Igor Jahatspanian

    Halitosis is a condition associated with bad breath. Although halitosis is a disease in its own right, it is often a symptom of more serious diseases (diabetes mellitus, renal failure, azotemia, etc.). The currently used method for diagnosing halitosis is the organoleptic method, which relies on a trained specialist evaluating the patient's breath odor. This approach to diagnosing halitosis is subjective, uncomfortable for both patient and doctor, and necessitates the involvement of a specially trained professional. As an alternative, instrumental diagnostics employing metal oxide semiconductor (MOS) sensor arrays offer a promising avenue by enabling patient classification through predeveloped models. This paper considers the application of seven MOS sensors of different compositions at three different temperatures. Different methods of chemometric data analysis were applied: k‐nearest neighbors (kNN), decision trees (DT), support vector machine (SVM), logistic regression (LR), and projection on latent structures discrimination analysis (PLSDA). All applied methods demonstrated their effectiveness and achieved selectivity, sensitivity, and accuracy values exceeding 85%. Additionally, a combined classifier leveraging responses from all previously studied classifiers was explored, achieving near‐perfect classification accuracy.