Technology for Objective Control of Functional Capacities and Stress of Military Servicemen Based on In-Depth ECG Analysis and Machine Learning

• Published in: Kìbernetika ta komp'ûternì tehnologìï (Online) no. 3; pp. 107 - 117
• Main Authors: Chaikovsky, Illya, Tulchinsky, Vadim, Golovynskyi, Andrii, Bocharov, Mykhailo, Kravchenko, Nataliia, Parovska, Sofia, Ryzhenko, Tetiana
• Format: Journal Article
• Language: English
• Published: V.M. Glushkov Institute of Cybernetics 29.09.2025
• Subjects: combat stress; ecg analysis; functional state of soldiers; machine learning; medical data analysis; objective assessment of condition; psychophysiological state of soldiers
• ISSN: 2707-4501; 2707-451X; 2707-451X
• DOI: 10.34229/2707-451X.25.3.10

Introduction. Analysis of changes in armed struggle in Russia's war against Ukraine in 2024 (Assessment of Russia's offensive campaign, December 20, 2024 | Institute for the Study of War, The first robotic operation on the battlefield: ISW noted the technological progress of the Armed Forces of Ukraine; War in Ukraine – how many Ukrainian and Russian military personnel died - UNIAN) indicates that the aggressor is maintaining a strategy of attrition, an increase in the number of groups, the intensity of fire exposure using the massive use of reconnaissance and strike high-precision weapons, a significant increase in the vulnerability of weapons, equipment and military personnel, an increase in sanitary and psychogenic losses, an increase in the need for mobilization resources, a decrease in the professional and physical qualities of the mobilizer, and increased requirements for training to perform combat missions in real time. The purpose of the paper is to create a technology for objective assessment and prediction of the functional state and stress resistance of individual servicemen and entire units, simple and suitable for use without special training near the front line, in training camps, etc. Results. A functional state model based on 26,976 ECG from military and civilians is proposed. Based on 23 parameters and t-SNE/UMAP transformations, 2 clusters were identified. Members of the first one are significantly more likely to suffer from high stress and deterioration of functional state when the second cluster is related to higher stress resistance. The model provides a generalized express assessment and prediction of the functional state psycho-emotional and resource-energy components as well as a more accurate assessment calculated according to the presented algorithm. The model implementation is based on using small portable ECG devices able to register 1-channel ECG signal from fingers without the need to undress and attach electrodes. Simplicity of use is important in the field and front conditions. Conclusions. The developed technology for predicting stress and fatigue resistance based on simple, rapid objective measurement is already being used in practice and is receiving positive feedback in units of various specializations. The ability to predict stress resistance based on objective measurements does not replace the need for a professional psychologist to treat stress. However, it can become an additional simple and powerful tool for better allocation of human resources and forecasting the future combat capabilities of units. It is important that the model can be used already in the training camp, that is, before the servicemen receive real combat experience. Keywords: functional state of soldiers, combat stress, psychophysiological state of soldiers, objective assessment of condition, ECG analysis, machine learning, medical data analysis.