Smartwatch gait coordination index: New measure for human gait utilizing smartwatch sensor

Abstract

Human walking reflects the state of human health. Numerous medical studies have been conducted to analyze walking patterns and to diagnose disease progression. However, this process requires expensive equipment and considerable time and manpower. Smartwatches are equipped with gyro sensors to detect human movements and graph-walking patterns. To measure the abnormality in walking using this graph, we developed a smartwatch gait coordination index (SGCI) and examined its usefulness. The phase coordination index was applied to analyze arm movements. Based on previous studies, the phase coordination index formula was applied to graphs obtained from arm movements, showing that arm and leg movements during walking are correlated with each other. To prove this, a smartwatch was worn on the arms and legs of 8 healthy adults and the difference in arm movements was measured. The SGCI values with abnormal walking patterns were compared with the SGCI values obtained during normal walking. In the first experiment, the measured leg SGCI in normal walking averaged 9.002 ± 3.872 and the arm SGCI averaged 9.847 ± 6.115. The movements of both arms and legs showed stable sinusoidal waves. In fact, as a result of performing a paired t test of both exercise phases measured by the strike point using the maximum and minimum values, it was confirmed that the 2 exercises were not statistically different, as it yielded a P value of 0.469 (significance level α = 0.05). The arm SGCI measured after applying the 3 kg weight impairment on 1 leg was 22.167 ± 4.705. It was confirmed that the leg SGCI and 3 kg weight arm SGCI were statistically significant, as it yielded a P value of 0.001 (significance level α = 0.05). The SCGI can be automatically and continuously measured with the gyro sensor of the smartwatch and can be used as an indirect indicator of human walking conditions.