Assessing Gross Motor and Gait Function Using Hip–Knee Cyclograms in Ambulatory Children with Spastic Cerebral Palsy

TL;DR

This study shows that hip–knee cyclogram metrics can serve as digital biomarkers to assess motor and gait function in children with cerebral palsy, offering a simpler alternative to traditional hospital-based evaluations.

Contribution

The study introduces hip–knee cyclogram parameters as accessible digital biomarkers for evaluating motor and gait function in children with spastic CP.

Findings

Hip–knee cyclogram parameters showed moderately strong correlations with GMFM-66, gait speed, GDI, and GPS subscores.

Swing phase area was the strongest correlate and used to estimate GMFM-66 and gait speed with R² values of 0.301 and 0.484, respectively.

The PC1/PC2 ratio correlated with selective motor control as measured by the SCALE (R² = 0.320).

Abstract

Children with cerebral palsy (CP) often experience weakness, spasticity, and muscle shortening, leading to deficits in gross motor, gait, and selective motor functions. GMFM-66, gait analysis, and the SCALE are commonly used in this regard, but are limited by time and space constraints in hospital settings. This study aims to identify digital biomarkers to assess gross motor, gait, and selective motor functions in children with CP, using simple kinematic data for hip–knee cyclogram-based analysis.

What are the main findings?
The parameters of the hip–knee cyclogram showed moderately strong correlations with GMFM-66, gait speed, GDI, and the sagittal plane subscore of the GPS for the hip and knee, while the principal component ratio correlated with the SCALE.In particular, the swing phase area showed the strongest correlation and was used to build regression models to estimate the GMFM-66 and gait speed, while the principal component ratio was used to assess the SCALE.

The parameters of the hip–knee cyclogram showed moderately strong correlations with GMFM-66, gait speed, GDI, and the sagittal plane subscore of the GPS for the hip and knee, while the principal component ratio correlated with the SCALE.

In particular, the swing phase area showed the strongest correlation and was used to build regression models to estimate the GMFM-66 and gait speed, while the principal component ratio was used to assess the SCALE.

What are the implications of the main finding?
Cyclogram metrics showed potential as digital biomarkers for assessing gross motor, gait, and selective motor functions in children with cerebral palsy.

Cyclogram metrics showed potential as digital biomarkers for assessing gross motor, gait, and selective motor functions in children with cerebral palsy.

Weakness, spasticity, and muscle shortening are common in children with cerebral palsy (CP), leading to deficits in gross motor, gait, and selective motor functions. While traditional assessments, such as the Gross Motor Function Measure (GMFM-66), instrumented gait analysis, and the Selective Control Assessment of the Lower Extremity (SCALE), are widely used, they are often limited by the resource-intensive nature of hospital-based evaluations. We employed cyclogram-based analysis, utilizing simple hip and knee joint kinematics to assess clinical measures, including GMFM-66, normalized gait speed, the gait deviation index (GDI), and the gait profile score (GPS). Principal component analysis was used to quantify the cyclogram shape characteristics. A total of 144 children with ambulatory spastic CP were included in the study. All the cyclogram parameters were significantly correlated with GMFM-66, gait speed, the GDI, and the sagittal plane subscore of the GPS for the hip and knee, with the swing phase area showing the strongest correlation. Regression models based on the swing phase area were used to estimate the GMFM-66 (R2 = 0.301) and gait speed (R2 = 0.484). The PC1/PC2 ratio showed a moderate correlation with selective motor control, as measured by the SCALE (R2 = 0.320). These findings highlight the potential of hip–knee cyclogram parameters to be used as accessible digital biomarkers for evaluating motor control and gait function in children with bilateral spastic CP. Further prospective studies using wearable sensors, such as inertial measurement units, are warranted to validate and build upon these results.

Figures