Comparison of different machine learning models to enhance sacral acceleration-based estimations of running stride temporal variables and peak vertical ground reaction force

Abstract

Machine learning (ML) was used to predict contact (<mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:mrow><mml:msub><mml:mi>t</mml:mi><mml:mi>c</mml:mi></mml:msub></mml:mrow></mml:math>) and flight (<mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:mrow><mml:msub><mml:mi>t</mml:mi><mml:mi>f</mml:mi></mml:msub></mml:mrow></mml:math>) time, duty factor (DF) and peak vertical force (<mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:mrow><mml:msub><mml:mi>F</mml:mi><mml:mrow><mml:mi>v</mml:mi><mml:mo>,</mml:mo><mml:mrow><mml:mrow><mml:mi>max</mml:mi></mml:mrow></mml:mrow></mml:mrow></mml:msub></mml:mrow></mml:math>) from IMU-based estimations. One hundred runners ran on an instrumented treadmill (9-13 km/h) while wearing a sacral-mounted IMU. Linear regression (LR), support vector regression and two-layer neural-network were trained (80 participants) using IMU-based estimations, running speed, stride frequency and body mass. Predictions (remaining 20 participants) were compared to gold standard (kinetic data collected using the force plate) by calculating the mean absolute percentage error (MAPE). MAPEs of <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:mrow><mml:msub><mml:mi>F</mml:mi><mml:mrow><mml:mi>v</mml:mi><mml:mo>,</mml:mo><mml:mrow><mml:mrow><mml:mi>max</mml:mi></mml:mrow></mml:mrow></mml:mrow></mml:msub></mml:mrow></mml:math> did not significantly differ among its estimation and predictions (<i>P</i> = 0.37), while prediction MAPEs for <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:mrow><mml:msub><mml:mi>t</mml:mi><mml:mi>c</mml:mi></mml:msub></mml:mrow></mml:math>, <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:mrow><mml:msub><mml:mi>t</mml:mi><mml:mi>f</mml:mi></mml:msub></mml:mrow></mml:math> and DF were significantly smaller than corresponding estimation MAPEs (<i>P</i> ≤ 0.003). There were no significant differences among prediction MAPEs obtained from the three ML models (<i>P</i> ≥ 0.80). Errors of the ML models were equal to or smaller than (≤32%) the smallest real difference for the four variables, while errors of the estimations were not (15-45%), indicating that ML models were sufficiently accurate to detect a clinically important difference. The simplest ML model (LR) should be used to improve the accuracy of the IMU-based estimations. These improvements may be beneficial when monitoring running-related injury risk factors in real-world settings.