Dataset Information

Machine Learning to Quantify Physical Activity in Children with Cerebral Palsy: Comparison of Group, Group-Personalized, and Fully-Personalized Activity Classification Models.

ABSTRACT: Pattern recognition methodologies, such as those utilizing machine learning (ML) approaches, have the potential to improve the accuracy and versatility of accelerometer-based assessments of physical activity (PA). Children with cerebral palsy (CP) exhibit significant heterogeneity in relation to impairment and activity limitations; however, studies conducted to date have implemented "one-size fits all" group (G) models. Group-personalized (GP) models specific to the Gross Motor Function Classification (GMFCS) level and fully-personalized (FP) models trained on individual data may provide more accurate assessments of PA; however, these approaches have not been investigated in children with CP. In this study, 38 children classified at GMFCS I to III completed laboratory trials and a simulated free-living protocol while wearing an ActiGraph GT3X+ on the wrist, hip, and ankle. Activities were classified as sedentary, standing utilitarian movements, or walking. In the cross-validation, FP random forest classifiers (99.0-99.3%) exhibited a significantly higher accuracy than G (80.9-94.7%) and GP classifiers (78.7-94.1%), with the largest differential observed in children at GMFCS III. When evaluated under free-living conditions, all model types exhibited significant declines in accuracy, with FP models outperforming G and GP models in GMFCS levels I and II, but not III. Future studies should evaluate the comparative accuracy of personalized models trained on free-living accelerometer data.

SUBMITTER: Ahmadi MN

PROVIDER: S-EPMC7411900 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundCerebral palsy (CP) is a physical disability that affects movement and posture. Approximately 17 million people worldwide and 34,000 people in Australia are living with CP. In clinical and kinematic research, goniometers and inclinometers are the most commonly used clinical tools to measure joint angles and positions in children with CP.ObjectiveThis paper presents collaborative research between the School of Electrical Engineering, Computing and Mathematical Sciences at Curtin University and a team of clinicians in a multicenter randomized controlled trial involving children with CP. This study aims to develop a digital solution for mass data collection using inertial measurement units (IMUs) and the application of machine learning (ML) to classify the movement features associated with CP to determine the effectiveness of therapy. The results were calculated without the need to measure Euler, quaternion, and joint measurement calculation, reducing the time required to classify the data.MethodsCustom IMUs were developed to record the usual wrist movements of participants in 2 age groups. The first age group consisted of participants approaching 3 years of age, and the second age group consisted of participants approaching 15 years of age. Both groups consisted of participants with and without CP. The IMU data were used to calculate the joint angle of the wrist movement and determine the range of motion. A total of 9 different ML algorithms were used to classify the movement features associated with CP. This classification can also confirm if the current treatment (in this case, the use of wrist extension) is effective.ResultsUpon completion of the project, the wrist joint angle was successfully calculated and validated against Vicon motion capture. In addition, the CP movement was classified as a feature using ML on raw IMU data. The Random Forrest algorithm achieved the highest accuracy of 87.75% for the age range approaching 15 years, and C4.5 decision tree achieved the highest accuracy of 89.39% for the age range approaching 3 years.ConclusionsAnecdotal feedback from Minimising Impairment Trial researchers was positive about the potential for IMUs to contribute accurate data about active range of motion, especially in children, for whom goniometric methods are challenging. There may also be potential to use IMUs for continued monitoring of hand movements throughout the day.Trial registrationAustralian New Zealand Clinical Trials Registry (ANZCTR) ACTRN12614001276640, https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=367398; ANZCTR ACTRN12614001275651, https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=367422.

Project description:BackgroundCerebral palsy (CP) is the most common physical disability among children (2.5 to 3.6 cases per 1000 live births). Inadequate physical activity (PA) is a major problem effecting the health and well-being of children with CP. Practical, yet accurate measures of PA are needed to evaluate the effectiveness of surgical and therapy-based interventions to increase PA. Accelerometer-based motion sensors have become the standard for objectively measuring PA in children and adolescents; however, current methods for estimating physical activity intensity in children with CP are associated with significant error and may dramatically underestimate HPA in children with more severe mobility limitations. Machine learning (ML) models that first classify the PA type and then predict PA intensity or energy expenditure using activity specific regression equations may be more accurate than standalone regression models. However, the feasibility and validity of ML methods has not been explored in youth with CP. Therefore, the purpose of this study was to develop and test ML models for the automatic identification of PA type in ambulant children with CP.MethodsTwenty two children and adolescents (mean age: 12.8?±?2.9 y) with CP classified at GMFCS Levels I to III completed 7 activity trials while wearing an ActiGraph GT3X+ accelerometer on the hip and wrist. Trials were categorised as sedentary (SED), standing utilitarian movements (SUM), comfortable walking (CW), and brisk walking (BW). Random forest (RF), support vector machine (SVM), and binary decision tree (BDT) classifiers were trained with features extracted from the vector magnitude (VM) of the raw acceleration signal using 10?s non-overlapping windows. Performance was evaluated using leave-one-subject out cross validation.ResultsSVM (82.0-89.0%) and RF (82.6-88.8%) provided significantly better classification accuracy than BDT (76.1-86.2%). Hip (82.7-85.5%) and wrist (76.1-82.6%) classifiers exhibited comparable prediction accuracy, while the combined hip and wrist (86.2-89.0%) classifiers achieved the best overall performance. For all classifiers, recognition accuracy was excellent for SED (94.1-97.9%), good to excellent for SUM (74.0-96.6%) and brisk walking (71.5-86.0%), and modest for comfortable walking (47.6-70.4%). When comfortable and brisk walking were combined into a single walking class, recognition accuracy ranged from 90.3 to 96.5%.ConclusionsML methods provided acceptable classification accuracy for detection of a range of activities commonly performed by ambulatory children with CP. The resultant models can help clinicians more effectively monitor bouts of brisk walking in the community. The results indicate that 2-step models that first classify PA type and then predict energy expenditure using activity specific regression equations are worthy of exploration in this patient group.

Project description:Background: Research in high income countries shows that people with cerebral palsy (CP) are less physically active than typically developing (TD) peers, but less is known regarding physical activity (PA) in those with CP in low-to-middle income countries. The aim of this study was to determine daily step count and levels of PA in adolescents and adults with CP living in urban South Africa, compared to TD peers, and to determine associations with sex, Gross Motor Function Classification System (GMFCS) level, body mass index and socio-economic status. Materials and Methods: This case-control study included 26 adolescents and 22 adults with CP (GMFCS Level I-V) and matched TD peers (25 and 30, respectively). Participants wore an ActiGraph GT3X for 7 consecutive days to determine step count and time (minutes per hour) spent in PA levels, including sedentary (SED), low physical activity (LPA) and moderate to vigorous physical activity (MVPA). Results: The daily step count and PA levels for ambulant adolescents with CP (GMFCS level I-III) were similar to TD peers, while this was less for adolescents classified in GMFCS level IV-V. Daily step count, SED and MVPA were similar for adults classified in GMFCS level I-II compared to TD adults, while all parameters were lower for adults using assistive devices (GMFCS level III) and non-ambulant adults (GMFCS level IV-V) compared to TD peers. Daily step count and PA levels were inversely associated with GMFCS, while no other associations were found. Conclusion: People with CP who were more mobile dependent (higher GMFCS level) were less active. However, adolescents and adults with CP classified as GMFCS levels I-II living in urban South Africa recorded similar step count and PA levels as their TD peers. This was also the case for adolescents using assistive devices, though not for those in the adult group (GMFCS level III). Furthermore, it was apparent that even the ambulant individuals with CP and TD cohorts were relatively inactive. Intervention programs for CP and TD adolescents should be aimed at finding strategies to keep adolescents physically active well into adulthood, in order to promote physical health, social and emotional well-being and independence.

Project description:IntroductionChildren and adolescents with cerebral palsy may be trapped in a vicious circle of low physical fitness, resulting in deconditioning that causes a further decrease in physical activity (PA), a lower quality of life and an increased risk of developing non-communicable diseases. Therefore, establishing a healthy and active lifestyle during childhood is even more important for individuals with a disability. However, the factors that influence habitual PA in children and adolescents with cerebral palsy remain unknown.The present protocol outlines a prospective cohort study with the aim of investigating potential predictors of habitual PA in children and adolescents with cerebral palsy in order to provide evidence for optimising PA levels and associated overall health.Methods and analysisThis prospective cohort study will enrol participants with cerebral palsy between the ages of 8 and 15 years at Gross Motor Function Classification System levels I-III. Using a modified version of the International Classification of Functioning, Disability and Health model as a conceptual analytical framework, the analysis will be divided into six components and will provide predictors for habitual PA measured by accelerometry. The potential predictive variables are registry data on physical function (Danish Cerebral Palsy Follow-Up Programme); validated proxy-reported questionnaires on quality of life (Paediatric Quality of Life Inventory), overall health, pain and participation in daily activities (Paediatric Outcomes Data Collection Instrument) and supplementary questions regarding sleep, screen time and socioeconomic status.Ethics and disseminationThe project is approved by the Danish Data Protection Agency (19/16396) and has been declared not notifiable by the Regional Committee on Health Research Ethics, cf. Committee Act Art. 14, paragraph 1 (S-20192000-23). The study results will be published in international peer-reviewed journals, presented at international conferences, and published in a PhD dissertation.Trial registration numberNCT04614207.

Project description:Accelerometers are widely used to measure physical activity behaviour, including in children. The traditional method for processing acceleration data uses cut points to define physical activity intensity, relying on calibration studies that relate the magnitude of acceleration to energy expenditure. However, these relationships do not generalise across diverse populations and hence they must be parametrised for each subpopulation (e.g., age groups) which is costly and makes studies across diverse populations and over time difficult. A data-driven approach that allows physical activity intensity states to emerge from the data, without relying on parameters derived from external populations, offers a new perspective on this problem and potentially improved results. We applied an unsupervised machine learning approach, namely a hidden semi-Markov model, to segment and cluster the raw accelerometer data recorded (using a waist-worn ActiGraph GT3X+) from 279 children (9-38 months old) with a diverse range of developmental abilities (measured using the Paediatric Evaluation of Disability Inventory-Computer Adaptive Testing measure). We benchmarked this analysis with the cut points approach, calculated using thresholds from the literature which had been validated using the same device and for a population which most closely matched ours. Time spent active as measured by this unsupervised approach correlated more strongly with PEDI-CAT measures of the child's mobility (R2: 0.51 vs 0.39), social-cognitive capacity (R2: 0.32 vs 0.20), responsibility (R2: 0.21 vs 0.13), daily activity (R2: 0.35 vs 0.24), and age (R2: 0.15 vs 0.1) than that measured using the cut points approach. Unsupervised machine learning offers the potential to provide a more sensitive, appropriate, and cost-effective approach to quantifying physical activity behaviour in diverse populations, compared to the current cut points approach. This, in turn, supports research that is more inclusive of diverse or rapidly changing populations.

Project description:BackgroundAdolescents with cerebral palsy (CP) who have mobility limitations have almost no access to inexpensive and enjoyable home-based programs that can be disseminated on a large scale to help them independently manage their health through participation in leisure-time physical activity (LTPA).ObjectiveThe primary aim of this study was to determine the preliminary efficacy of the early adoption phase of an adult Movement-to-Music (M2M) program with behavioral telecoaching for increasing LTPA and activity participation compared with a waitlist control group in adolescents with CP. The secondary aim was to explore the effects of the program on perceived levels of pain and fatigue. The tertiary aim was to qualitatively evaluate the factors that influenced adherence and develop a theory that would inform the development of a more targeted M2M telehealth program for this group.MethodsThis randomized controlled trial piloted a 4-week M2M program with weekly behavioral telecoaching among 58 adolescents with CP who walked or used wheelchairs. The participants were randomized into one of 2 groups: M2M or control, which maintained their daily activities. M2M included videos that participants were asked to complete 3 times each week at home (asynchronous training). Adherence to video minutes was objectively measured using cloud-based analytics. Changes in activity and LTPA participation were measured before and after the intervention using the Children's Assessment of Participation and Enjoyment total domain scores and active physical recreation domain scores, respectively. Perceived pain and fatigue were measured using the National Institutes of Health Neuro-QoL short forms. The changes in scores were compared between the groups using analysis of covariance. A grounded theory approach was used to analyze one-on-one interviews, coaching notes, and feedback surveys.ResultsA total of 58 people were enrolled, of which 49 (84%) completed the primary outcome follow-up assessment. The mean adherence to the prescribed exercise video minutes across all 4 weeks was 68%, starting from 90% in week 1 and gradually declining to 43% in week 4. Mean adherence to coaching calls was 91%. Analysis of covariance revealed a statistically significant difference between the pre- to postchange scores for Children's Assessment of Participation and Enjoyment Active Physical Recreation-Intensity domain scores in favor of the intervention group (F1,47=8.76; P=.005; effect size=0.17, also known as volume of LTPA). The qualitative findings highlighted 5 critical factors that influenced participants' adherence to the program: caregiver support, video elements, suitable exercises, music, and behavioral coaching.ConclusionsThis project determined that adolescents with CP responded well to an M2M telehealth program that could enhance their LTPA levels. This paper describes a theory in which adherence to a telehealth LTPA program can be optimized through functional and age-specific modifications for adolescents with CP.Trial registrationClinicalTrials.gov NCT04264390; https://clinicaltrials.gov/ct2/show/NCT04264390.

Dataset Information

Machine Learning to Quantify Physical Activity in Children with Cerebral Palsy: Comparison of Group, Group-Personalized, and Fully-Personalized Activity Classification Models.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets