본문 바로가기
  • Home

An artificial intelligence-based predictive model for periodontal disease expenditures using community-level health environmental factors from public data

  • Journal of Korean society of Dental Hygiene
  • Abbr : J Korean Soc Dent Hyg
  • 2026, 26(2), pp.243~252
  • DOI : 10.13065/jksdh.2026.26.2.11
  • Publisher : Korean Society of Dental Hygiene
  • Research Area : Medicine and Pharmacy > Dentistry
  • Received : February 24, 2026
  • Accepted : April 8, 2026
  • Published : April 30, 2026

Kyu-Seok Kim 1 Narae Jung 2 Jung Yun Kang 3

1한국폴리텍대학 인공지능소프트웨어과
2한서대학교 치위생학과
3연세대학교 치위생학과

Accredited

ABSTRACT

Objectives: This study aimed to develop and evaluate a deep neural network (DNN)-based predictive model for periodontal disease expenditures using community-level health environmental factors derived from public data. Methods: A total of 1,020 monthly records from 17 regions between January 2020 and December 2024 were analyzed. Independent variables included health behaviors, demographic characteristics, socioeconomic factors, and healthcare resource accessibility. A DNN model was constructed and evaluated using the mean absolute percentage error (MAPE), while permutation feature importance (PFI) was applied to quantify the relative contribution of each variable. Results: The DNN model achieved a mean MAPE of 11.01% (range: 9.35–12.51; SD: 0.83) across 10 repeated trials, indicating good predictive performance according to the Lewis (1982) criteria. PFI analysis identified total population, proportion of single-person households, and gender ratio as the most influential predictors of periodontal disease expenditures. Conclusions: These findings suggest that periodontal disease expenditures are shaped by complex interactions among demographic, socioeconomic, and behavioral factors, which can be effectively captured by AI-based predictive models. This study provides preliminary evidence that healthcare resource allocation and oral health policy development can benefit from AI-based approaches utilizing publicly available data.

Journal Copyright Policy

BY-NC 아이콘

This license enables reusers to distribute, remix, adapt, and build upon the material in any medium or format for noncommercial purposes only, and only so long as attribution is given to the creator.

Citation status

* References for papers published after 2025 are currently being built.