VO2 estimation using 6-axis motion sensing data
Masato Miyatake, Naoteru Nakamura, Takashi Nagata, Akira Yuuki, Hiroyuki Yomo, Takashi Kawabata, Shinsuke Hara · 2016
The information on energy expenditure of people playing various types of sports can be exploited for diagnosing their physical conditions, which helps the avoidance of their diseases and injuries during sports activities. The energy expenditure can be calculated from oxygen consumption (VO2) that can be directly measured by a VO2meter, however, it is not reasonable for people playing sports to wear it during intensive activities. Therefore, VO2estimation using acceleration data obtained by small accelerometer has been proposed. In this paper, we first investigate the accuracy of VO2estimation based on l1and l2norms calculated from triaxial acceleration data. Then, we focus on the estimation using motion sensor that can measure 6-axis motion data. The estimation accuracy is investigated based on experimental data consisting of VO2and motion sensing data of people with different intensities of exercises, which were measured by a VO2meter and 6-axis motion sensor, respectively. Our numerical results show that the proposed estimation method, which treats data on 6 different axes separately and independently, can improve the estimation accuracy in comparison to a reference method that is based on the estimation using norm information.