This paper studies how to describe, using a piecewise linear dynamical model, the short-term effects of fatigue and recovery on the strength of pelvic floor muscles. Specifically, we first adapt a known model that describes short-term fatigue in skeletal muscles to the specific problem of describing fatigue in pelvic floor muscles when performing Kegel exercises, and then propose a strategy to learn the modelŠs parameters from field data.In details, we estimate the model parameters using a least squares approach starting from measurement data that has been obtained from three healthy women using a dedicated vaginal pressure sensor array and a connected mobile app which gamifies the Kegel exercising experience.We show that describing the pelvic floor muscles behaviour in terms of short-term fatigue and recovery factors plus learning the associated parameters from data from healthy women leads to the possibility of precisely forecasting how much pressure the players will exert while playing the game.By cross-learning and cross-testing individual models from the three volunteers we also discover that the models need to be individualized: indeed, the numerical results indicate that, generically, using data from one player to model another leads to potentially drastically lower forecasting capabilities.
Data-driven modelling of fatigue in pelvic floor muscles when performing Kegel exercises
Varagnolo D.;
2019
Abstract
This paper studies how to describe, using a piecewise linear dynamical model, the short-term effects of fatigue and recovery on the strength of pelvic floor muscles. Specifically, we first adapt a known model that describes short-term fatigue in skeletal muscles to the specific problem of describing fatigue in pelvic floor muscles when performing Kegel exercises, and then propose a strategy to learn the modelŠs parameters from field data.In details, we estimate the model parameters using a least squares approach starting from measurement data that has been obtained from three healthy women using a dedicated vaginal pressure sensor array and a connected mobile app which gamifies the Kegel exercising experience.We show that describing the pelvic floor muscles behaviour in terms of short-term fatigue and recovery factors plus learning the associated parameters from data from healthy women leads to the possibility of precisely forecasting how much pressure the players will exert while playing the game.By cross-learning and cross-testing individual models from the three volunteers we also discover that the models need to be individualized: indeed, the numerical results indicate that, generically, using data from one player to model another leads to potentially drastically lower forecasting capabilities.Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.