Ines Vandekerckhove, Dhruv Gupta, Lars D'Hondt, Marleen Van den Hauwe, Anja Van Campenhout, Liesbeth De Waele, Nathalie Goemans, Kaat Desloovere, Friedl De Groote
{"title":"Predictive simulations of common gait features in children with Duchenne muscular dystrophy","authors":"Ines Vandekerckhove, Dhruv Gupta, Lars D'Hondt, Marleen Van den Hauwe, Anja Van Campenhout, Liesbeth De Waele, Nathalie Goemans, Kaat Desloovere, Friedl De Groote","doi":"10.1016/j.gaitpost.2023.07.257","DOIUrl":null,"url":null,"abstract":"Predictive simulations of gait can improve our understanding of how underlying impairments contribute to gait pathology in children with Duchenne muscular dystrophy (DMD). This is essential to make progress in gait rehabilitation and orthotic treatments aiming at prolonging ambulation in DMD. Yet, there is still a need to evaluate if predictive simulations can capture the key features of DMD gait. Can we simulate DMD gait pathology? 3D gait analysis was collected in three boys with DMD, who were situated at different stages of the disease progression. Muscle weakness was measured using a fixed dynamometer [1]. Muscle stiffness and contractures were assessed using goniometry and clinical scales. For each subject, a generic musculoskeletal model [2] was scaled to the subject’s anthropometry based on marker data. The maximal isometric muscle forces (MIMF), joint stiffness, properties of the foot-ground contact model, weights of the cost function and imposed walking speed were scaled to reflect the child’s dimensions. Subject-specific muscle weakness was modeled by decreasing active MIMF based on the individual’s weakness scores. Muscle stiffness and contractures were modeled by shifting and increasing the steepness of the passive force-length relationship of the assessed muscles. Gait was predicted by minimizing a cost function while imposing the gait speed and periodicity of the gait pattern (without relying on motion capture data) [3]. For each subject, simulations were performed based on four models: (1) reference (child’s dimensions), (2) weakness, (3) stiffness, and (4) combination of weakness and stiffness. Root mean squared error (RMSE) between the simulated kinematics and the mean experimental kinematics was calculated. Fig. 1 shows the experimental data and simulation results of DMD1 (10.6years), DMD2 (15.6years) and DMD3 (11.1years). The predicted gait patterns are closer to the experimental data when modeling weakness and stiffness. The sum of RMSEs between predicted and experimental kinematics decreased from 40.9 to 36.2 between model1 to model4 for DMD1, from 47.5 to 30.3 for DMD2 and from 48.2 to 39.2 for DMD3. The increasing gait pathology over the three cases with increasing severity of muscle impairments, was also reflected in the predictive simulations.Download : Download high-res image (273KB)Download : Download full-size image Several key features of the DMD gait, such as tiptoeing gait, increased anterior pelvic tilt, reduced knee flexion during stance and drop foot in swing, were reasonably captured in the predictive simulations. However, the exaggerated lumbar extension was not fully captured. Differences between simulations and experiments might be due to the use of a simple trunk model. In addition, foot deformities were not yet modeled. In the future, we will further refine the model and personalization workflow by using data from instrumented stiffness assessment. Nevertheless, the current results show the potential of predictive simulations to improve our insights in the progressive gait pathology in boys with DMD.","PeriodicalId":94018,"journal":{"name":"Gait & posture","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2023-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Gait & posture","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1016/j.gaitpost.2023.07.257","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Predictive simulations of gait can improve our understanding of how underlying impairments contribute to gait pathology in children with Duchenne muscular dystrophy (DMD). This is essential to make progress in gait rehabilitation and orthotic treatments aiming at prolonging ambulation in DMD. Yet, there is still a need to evaluate if predictive simulations can capture the key features of DMD gait. Can we simulate DMD gait pathology? 3D gait analysis was collected in three boys with DMD, who were situated at different stages of the disease progression. Muscle weakness was measured using a fixed dynamometer [1]. Muscle stiffness and contractures were assessed using goniometry and clinical scales. For each subject, a generic musculoskeletal model [2] was scaled to the subject’s anthropometry based on marker data. The maximal isometric muscle forces (MIMF), joint stiffness, properties of the foot-ground contact model, weights of the cost function and imposed walking speed were scaled to reflect the child’s dimensions. Subject-specific muscle weakness was modeled by decreasing active MIMF based on the individual’s weakness scores. Muscle stiffness and contractures were modeled by shifting and increasing the steepness of the passive force-length relationship of the assessed muscles. Gait was predicted by minimizing a cost function while imposing the gait speed and periodicity of the gait pattern (without relying on motion capture data) [3]. For each subject, simulations were performed based on four models: (1) reference (child’s dimensions), (2) weakness, (3) stiffness, and (4) combination of weakness and stiffness. Root mean squared error (RMSE) between the simulated kinematics and the mean experimental kinematics was calculated. Fig. 1 shows the experimental data and simulation results of DMD1 (10.6years), DMD2 (15.6years) and DMD3 (11.1years). The predicted gait patterns are closer to the experimental data when modeling weakness and stiffness. The sum of RMSEs between predicted and experimental kinematics decreased from 40.9 to 36.2 between model1 to model4 for DMD1, from 47.5 to 30.3 for DMD2 and from 48.2 to 39.2 for DMD3. The increasing gait pathology over the three cases with increasing severity of muscle impairments, was also reflected in the predictive simulations.Download : Download high-res image (273KB)Download : Download full-size image Several key features of the DMD gait, such as tiptoeing gait, increased anterior pelvic tilt, reduced knee flexion during stance and drop foot in swing, were reasonably captured in the predictive simulations. However, the exaggerated lumbar extension was not fully captured. Differences between simulations and experiments might be due to the use of a simple trunk model. In addition, foot deformities were not yet modeled. In the future, we will further refine the model and personalization workflow by using data from instrumented stiffness assessment. Nevertheless, the current results show the potential of predictive simulations to improve our insights in the progressive gait pathology in boys with DMD.