Motor And Balance Analysis Of Knee Osteoarthritis In Vivo

Background and objectivesThe methods currently used for knee osteoarthritis(KOA)assessment have certain limitations.A portable and reliable measurement is still required for the objective evaluation of KOA.Previous studies have shown the possibility of quantifying motor and balance performance of KOA.The Functional Movement Screen(FMS)is a screening system that requires a balance of mobility and stability.But it has not been used in KOA assessment,and it is unclear whether it is more sensitive than daily tasks such as walking and sit-to-stand in patients with KOA.The markerless motion capture system is suitable for objective functional monitoring of patients inside or outside the clinical setting.Nevertheless,whether markerless motion capture technology can detect abnormal motor performance in patients with KOA has not been explored,and its detection validity needs to be further verified.Besides,ultrasound imaging technology has the characteristics of high real-time performance,low cost,and no radiation.It is unclear whether an in vivo detection system of the tibiofemoral joint can be built using robotic arm ultrasonic imaging combined with registration of CT images to ultrasound images technology and applied to the evaluation of KOA.As a result,the study’s objectives were(ⅰ)to build a detection system of tibiofemoral joint in vivo based on robotic arm ultrasonic imaging combined with registration of CT images to ultrasound images technology and evaluate the accuracy of the system,(ⅱ)to validate the markerless motion capture system by comparing the kinematic parameters of interest measured by the markerless motion capture system to those measured by the optical motion capture system,and(ⅲ)to evaluate which kinematic parameters differentiated between patients with different Kellgren-Lawrence grades of KOA and healthy controls during FMS(deep squat,hurdle step,and in-line lunge)and daily activities(walking,sit-to-stand)based on the markerless motion capture system.Methods and materials1.CT and ultrasound images of pork tibiofemoral joints and femur were collected and registered after three-dimensional reconstruction.After femur’s movement of six degrees of freedom,ultrasound images were collected and reconstructed and then registered to CT three-dimensional model.The displacement between CT threedimensional model and real displacement were compared to calculate registration accuracy.2.15 healthy volunteers were recruited.All participants completed five tasks,including the deep squat,hurdle step,in-line lunge,walking,and sit-to-stand.All tasks were simultaneously recorded by the markerless motion capture system and the optical motion capture system.The normalized waveforms of interested kinematic parameters from both systems were compared using the coefficient of multiple correlation(CMC)and the root mean squared error(RMSE).3.60 patients with KOA and healthy controls(HC)participated in the study.All participants carried out the same five tasks as those in the validity verification experiment.All tasks were recorded by the markerless motion capture system only.The entire kinematic waveforms and discrete kinematic parameters were compared.Results1.Using three-dimensional geometric information for ultrasound-CT manual registration to build a detection system of tibiofemoral joint in vivo is feasible.Translation accuracy ranges from 0.94mm to 1.16mm,and rotation accuracy ranges from 0.63° to 1.50°.2.The CMCs of all parameters were good to excellent,except for the weak CMCs of trunk frontal angle during squatting,hurdle step,and sit-to-stand.The RMSEs of the trunk angle and medial-lateral displacement of the COM(COM ML displacement)were low,whereas the RMSEs of the knee and hip flexion and extension were relatively high.3.Statistical parametric or nonparametric mapping analysis revealed differences in FMS but not in walking between patients and HC as well as patients and patients.Discrete kinematic parameter analysis revealed differences between patients and HC,as well as between patients and patients,during both FMS and daily activities.The discriminant indexes were the knee flexion and extension waveform and peak knee flexion angle during the deep squat,the knee and hip flexion and extension waveform during the hurdle step,the knee and hip flexion and extension waveform,trunk sagittal and frontal angle waveform,peak knee flexion angle,peak trunk sagittal and frontal angle,and peak COM ML displacement during the in-line lunge,the peak trunk frontal angle during the sit-to-stand,the peak hip extension angle and step length during the walking.Conclusions1.The ultrasound-CT registration method proposed in this article is feasible,and the accuracy can meet the requirements of tibiofemoral dynamic detection.Since the methods of ultrasound images’ obtaining by robotic arm and registration need to be further improved,the system cannot meet clinical application requirements in the short term.2.The markerless motion capture system is an alternative for evaluating motor and balance performances outside of the laboratory environment.3.Both FMS and daily movements can detect motor and balance abnormalities in KOA patients,with FMS being more sensitive.The knee flexion angle,trunk sagittal angle,and trunk frontal angle during the in-line lunge could be candidate indicators for the objective and quantifiable evaluation of KOA.