Development of a Mathematical Model for Personalized Estimation of Life Expectancy in Ukraine

• Published in: Kìbernetika ta komp'ûternì tehnologìï (Online) no. 2; pp. 47 - 60
• Main Authors: Malanin, Vladyslav, Chaykovsky, Illya
• Format: Journal Article
• Language: English
• Published: V.M. Glushkov Institute of Cybernetics 06.06.2025
• Subjects: life expectancy; mathematical model; medical cybernetics; personalized prediction; ukraine
• ISSN: 2707-4501; 2707-451X; 2707-451X
• DOI: 10.34229/2707-451X.25.2.4

Introduction. The issue of personalized life expectancy prediction is a relevant task in contemporary medical cybernetics, significantly impacting public health and social planning. It gains special importance during crisis situations such as war, the COVID-19 pandemic, economic hardships, and demographic changes that have intensified in Ukraine in recent years. Existing models for life expectancy prediction, such as the Mortality Population Risk Tool (MPoRT) and the Lee-Carter model, have significant limitations regarding the inclusion of local characteristics specific to the Ukrainian population. Thus, there is a necessity to develop adaptive and integrated models that account for specific demographic, medical, and socio-behavioral factors particular to Ukraine. This paper presents a mathematical model and the practical tool "Life Calculator," developed with the support of the Public Health Center of Ukraine's Ministry of Health. The model is based on a personalized approach to estimating an individual's survival probability, taking into account a broad spectrum of individual factors. The purpose of the paper is to develop an integrated methodology for personalized prediction of life expectancy considering unique individual factors of the Ukrainian population, and to create a practical software tool for its implementation. Results. A unique integrated methodology for predicting life expectancy has been proposed, incorporating 26 diverse impact factors classified into diagnostic (related to medical conditions) and non-diagnostic (behavioral, socio-economic). For the first time, factors highly relevant to current conditions in Ukraine, such as internally displaced person status, PTSD, and COVID-19 effects, have been integrated. A large-scale trial of the "Life Calculator" web tool confirmed substantial public interest in personalized predictions of this nature. Over the first six months of operation, more than 200,000 questionnaires were successfully completed, indicating the tool's ease of use and clarity. Statistical analysis revealed that the average expected life expectancy was 73 years for men and 83 years for women. The tool enables real-time generation of personalized survival tables, visually demonstrating survival probabilities up to the age of 110. This approach raises awareness of various factors' influence on life expectancy and motivates healthier lifestyle choices. Analysis also revealed the developed model's significant potential for integration with existing information systems and services such as Helsi and Diia, as well as the possibility of utilizing biometric data and data from wearable devices for faster data provision, more precise and detailed predictions. Conclusions. The proposed mathematical model and the web-based "Life Calculator" tool effectively address personalized life expectancy prediction for the Ukrainian population. Their widespread use confirms the high relevance and demand for such solutions. The model is adaptable to changing conditions and can serve as an important tool not only for personal health monitoring but also for developing effective public health and social planning strategies in Ukraine. Further improvement and integration of the model with other medical information resources will enhance prediction accuracy and broaden the practical application scope. Keywords: mathematical model, life expectancy, personalized prediction, medical cybernetics, Ukraine.