Energy efficiency in gait, activity, participation, and health status in children with cerebral palsy

The aim of the study was to establish if a relationship exists between the energy efficiency of gait, and measures of activity limitation, participation restriction, and health status in a representative sample of children with cerebral palsy (CP). Secondary aims were to investigate potential differences between clinical subtypes and gross motor classification, and to explore other relationships between the measures under investigation. A longitudinal study of a representative sample of 184 children with ambulant CP was conducted (112 males, 72 females; 94 had unilateral spastic C P, 84 had bilateral spastic C P, and six had non-spastic forms; age range 4-17y; Gross Motor Function Classification System Level I, n=57; Level II, n=91; Level III, n=22; and Level IV, n=14); energy efficiency (oxygen cost) during gait, activity limitation, participation restriction, and health status were recorded. Energy efficiency during gait was shown to correlate significantly with activity limitations; no relationship between energy efficiency during gait was found with either participation restriction or health status. With the exception of psychosocial health, all other measures showed significant differences by clinical subtype and gross motor classification. The energy efficiency of walking is not reflective of participation restriction or health status. Thus, therapies leading to improved energy efficiency may not necessarily lead to improved participation or general health.

Age-related changes in energy efficiency of gait, activity, and participation in children with cerebral palsy

Aim
The aim of this study was to use a prospective longitudinal study to describe age-related trends in energy efficiency during gait, activity, and participation in ambulatory children with cerebral palsy (CP).

Method
Gross Motor Function Measure (GMFM), Paediatric Evaluation of Disability Inventory (PEDI), and Lifestyle Assessment Questionnaire-Cerebral Palsy (LAQ-CP) scores, and energy efficiency (oxygen cost) during gait were assessed in representative sample of 184 children (112 male; 72 female; mean age 10y 9mo; range 4–16y) with CP. Ninety-four children had unilateral spastic CP, 84 bilateral spastic CP, and six had other forms of CP. Fifty-seven were classified as Gross Motor Function Classification System (GMFCS) level I, 91 as level II, 22 as level III, and 14 as level IV). Assessments were carried out on two occasions (visit 1 and visit 2) separated by an interval of 2 years and 7 months. A total of 157 participants returned for reassessment.

Results
Significant improvements in mean raw scores for GMFM, PEDI, and LAQ-CP were recorded; however, mean raw oxygen cost deteriorated over time. Age-related trends revealed gait to be most inefficient at the age of 12 years, but GMFM scores continued to improve until the age of 13 years, and two PEDI subscales to age 14 years, before deteriorating (p<0.05). Baseline score was consistently the single greatest predictor of visit 2 score. Substantial agreement in GMFCS ratings over time was achieved (?lw=0.74–0.76).

Interpretation
These findings have implications in terms of optimal provision and delivery of services for young people with CP to maximize physical capabilities and maintain functional skills into adulthood.

An efficient feature selection method for mobile devices with application to activity recognition

This paper presents a feature selection method for data classification, which combines a model-based variable selection technique and a fast two-stage subset selection algorithm. The relationship between a specified (and complete) set of candidate features and the class label is modelled using a non-linear full regression model which is linear-in-the-parameters. The performance of a sub-model measured by the sum of the squared-errors (SSE) is used to score the informativeness of the subset of features involved in the sub-model. The two-stage subset selection algorithm approaches a solution sub-model with the SSE being locally minimized. The features involved in the solution sub-model are selected as inputs to support vector machines (SVMs) for classification. The memory requirement of this algorithm is independent of the number of training patterns. This property makes this method suitable for applications executed in mobile devices where physical RAM memory is very limited. An application was developed for activity recognition, which implements the proposed feature selection algorithm and an SVM training procedure. Experiments are carried out with the application running on a PDA for human activity recognition using accelerometer data. A comparison with an information gain based feature selection method demonstrates the effectiveness and efficiency of the proposed algorithm.

Clustering-Based Ensemble Learning for Activity Recognition in Smart Homes

Application of sensor-based technology within activity monitoring systems is becoming a popular technique within the smart environment paradigm. Nevertheless, the use of such an approach generates complex constructs of data, which subsequently requires the use of intricate activity recognition techniques to automatically infer the underlying activity. This paper explores a cluster-based ensemble method as a new solution for the purposes of activity recognition within smart environments. With this approach activities are modelled as collections of clusters built on different subsets of features. A classification process is performed by assigning a new instance to its closest cluster from each collection. Two different sensor data representations have been investigated, namely numeric and binary. Following the evaluation of the proposed methodology it has been demonstrated that the cluster-based ensemble method can be successfully applied as a viable option for activity recognition. Results following exposure to data collected from a range of activities indicated that the ensemble method had the ability to perform with accuracies of 94.2% and 97.5% for numeric and binary data, respectively. These results outperformed a range of single classifiers considered as benchmarks.