121 A Newly-Developed Smart Insole System with Instant Reminder: Paves the Way towards Integrating Artificial Intelligence (AI) Technology to Improve Balance and Prevent Falls

Background Falls in senior people have high incidence& lead to severe injuries [1]. Application of smart wearable systems (with sensors to monitor user’s balance and corresponding instant reminder to let tusers adjust posture/motion) can effectively improve static standing balance [2], reduce reaction time and body sway in response to balance perturbation [3], improve walking pattern [4], and reduce the risk of falls [5, 6]. However, previous systems have not considered the daily monitor of user’s balance and falling risks, and the personalized reminder. Artificial intelligence (AI) and big data analytics have been widely used to monitor the daily physical activity [7], while few studies have utilized them to improve balance/gait and prevent falls. Methods This study has optimized previous devices by integrating AI technology and developed a new smart insole system. The system consisted of insoles with embedded sensors that can capture the foot motion and plantar pressure, smart watch that connected with insoles wirelessly and then transmitted the foot motion and force data to Cloud server via Wi-Fi, central Cloud server for big data transmission and storage, workstation for big data analytics and machine learning, and user interface for data visualization (e.g. smartphone, tablet, and/or laptop). Results & Discussion The system transmission rate was up to 30 Hz. The collected big data contained all sensor signals captured before and after delivering reminder, and from day-to-day monitoring of users. The customized reminder varied in the type, frequency, magnitude, and amount/dosage. This AI smart insole system enabled the monitor of daily balance and falling risks and the provision of timely-updated and customized reminder to users, which could potentially reduce the risk of falls and slips. It can also act as a balance-training device. References 1. Rubenstein.Age ageing, 2006. 35(suppl2):p.ii37-ii41. 2. Ma.Sensors, 2015. 15(12):p.31709-31722. 3. Ma&Lee.Human Movement Science, 2017. 55:p.54-60. 4. Ma.Topics in stroke rehabilitation, 2018. 25(1):p.20-27. 5. Wan.Archives of physical medicine& rehabilitation, 2016. 97(7):p.1210-1213. 6. Ma.Sensors, 2016. 16(4):p.434. 7. Badawi.Future Generation Computer Systems, 2017. 66:p.59-70.