Video game-based application for fall risk assessment: a proof-of-concept cohort study

Area of Science:

• Gerontology
• Biomedical Engineering
• Neurology

Background:

• Falls are a major cause of morbidity and mortality in the elderly, particularly those with polyneuropathy and cognitive decline.
• Conventional fall risk assessment tools have limited predictive value and fail to capture specific vulnerabilities.
• This study focuses on developing an innovative fall prediction tool for individuals with diabetes, a high-risk group.

Purpose of the Study:

• To develop and validate an engaging, video game-based fall risk assessment tool for individuals with diabetes.
• To compare the predictive accuracy of the novel tool against traditional clinical assessments.
• To identify key performance metrics within the video game that correlate with fall risk.

Main Methods:

• A proof-of-concept cohort study involving 152 participants with diabetes.
• Participants underwent traditional fall risk assessments (Timed Up and Go, Dynamic Gait Index, Berg Balance Scale).
• Participants also engaged in sensor-equipped insole-controlled video games assessing skillfulness, reaction time, sensation, endurance, balance, and muscle strength, in both seated and standing positions.

Main Results:

• Traditional assessments showed moderate accuracies (TUG: 58.7%, DGI: 58.3%, BBS: 47.5%).
• Video game-based assessments yielded significantly higher accuracies (82.8% seated, 88.6% standing).
• Key differentiating factors included endurance and balance, with AI analysis highlighting reaction times and pressure parameters as significant predictors.

Conclusions:

• Video game-based fall risk assessment surpasses traditional clinical tools in predictive accuracy for individuals with diabetes.
• This innovative approach offers a promising new resource for patient evaluation and fall prevention strategies.
• Further validation in larger, diverse cohorts is recommended to refine predictive capabilities in clinical settings.