Towards Collaboration with AI to Explore New Features of Dementia That Cannot Be Found from Medical Viewpoint

Authors

  • Keiki Takadama The University of Tokyo, Japan
  • Ryuki Ishizawa The University of Tokyo, Japan

DOI:

https://doi.org/10.1609/aaaiss.v8i1.42619

Abstract

This paper focuses on collaboration between human and AI and shows its potential of exploring new features that cannot be found by either human experts or AI. Concretely, this paper focuses on dementia and shows that new features of dementia based on heartrate cannot be found from the medical viewpoint but can be found from hints based on the analysis of AI. Through an analysis of heartrate of both dementia patients and healthy subjects, the following implications have been revealed: (1) an ability of dementia patients to maintain a stable heartrate as homeostasis is weaker than that of healthy subjects; and (2) the time-series features of heartrate obtained by AI (i.e., the statistical analysis and machine learning) is needed to be interpreted (by human) to find a potential symptom of dementia.

Downloads

Published

2026-05-18

How to Cite

Takadama, K., & Ishizawa, R. (2026). Towards Collaboration with AI to Explore New Features of Dementia That Cannot Be Found from Medical Viewpoint. Proceedings of the AAAI Symposium Series, 8(1), 767–771. https://doi.org/10.1609/aaaiss.v8i1.42619

Issue

Section

Will AI Light Up Human Creativity or Replace It?