Assessing cerebral palsy by direct measures of muscle dysfunction.
Muscle dysfunction in children with spastic cerebral palsy can be quantified by abnormal patterns of myoelectric activity, both within and between muscles.
Aim and background of project
Spastic cerebral palsy (CP) is a condition that results in abnormal muscle activation and movement patterns, and a reduction in the mobility of the individual. Current treatments include physiotherapy and surgical interventions to specific muscles. The choice of treatment may be informed by clinical assessments made at gait analysis laboratories, where the movement patterns and forces experienced by the joints are assessed. However, neither of these measures gives direct information about the abnormality and impairments in specific muscles. The goal of this project was to pioneer new methods that would directly quantify the dysfunction in individual muscles and thus, provide new information about the condition, and to inform the most appropriate treatments for the dysfunctional muscles in children with CP.
Electromyography (EMG) is the recording of the electrical signals the muscles emit when they are active. EMG signals are thus, direct measures of muscle activity, and it was proposed that these could be used to provide the required information on muscle dysfunction. The project investigated the EMG signals that have been routinely recorded during clinical gait assessment sessions at the Centre for Human Performance at the Royal National Orthopaedic Hospital, Stanmore. The project analysed the EMG signals from 17 children with spastic CP and 36 asymptomatic controls. The fine details within an EMG signal carry information about the physiology of the muscle. The project applied advanced signal processing methods to resolve these fine details; in particular, it used wavelet transformations to decompose the signals into their timefrequency properties, and then used a combination of correlation and principal component analysis to quantify the patterns of muscle firing.
Data Collection
Data for 17 children and young adults with spastic diplegic cerebral palsy (mean age 11.3 years, range 4-21).
The muscle strength for these children (CP) was assessed using the Oxford/ MRC scale from 5 for full strength to 0 for no contraction [11]: 4.06 ± 0.24 for the hip flexors and extensors, 3.84 ± 0.44 for the knee flexors, and extensors, and 3.09 ± 1.43 for the ankle dorsiflexors and plantarflexors. Six of the children with CP exhibited clonus at the ankle joint and 12 showed asymmetry in the strength and range of motion of between the left and right legs. These data had been previously collected as part of routine clinical gait analysis at the Motion Analysis Laboratory at the Royal National Orthopaedic Hospital, Stanmore, UK. The legal guardians gave their informed consent for the data being used for research purposes and this was retrospectively approved by the Central Office for Research Ethics Committee, Royal National Orthopaedic Hospital. Bilateral data for 36 healthy asymptomatic controls (mean age 10.8 years, range 3-21) was also analysed. Children walked 5-10 times at their self-selected comfortable speed along a 12 m walkway within the laboratory. Wherever possible, the children walked unaided, however, three children required the use of a Kayeframe to ambulate. Rest periods were given between trials as necessary in order to minimise fatigue. The 3D kinematics of the legs were determined (from the pelvis down) using a series of active infra-red markers (CODA mpx30 system: Charnwood Dynamics Ltd., Rothley, UK). Surface EMG was recorded bilaterally from the muscle bellies of the rectus femoris, biceps femoris, medial gastrocnemius, and tibialis anterior muscles. Bipolar electrodes (diameter and spacing; MA-300, Motion Lab Systems Inc., Baton Rouge, LA, USA) were adhered to the skin and the EMG data digitised on a 16-bit analog-to-digital converter and recorded at 2000 Hz.
Results
The results were all very encouraging, although in some places, surprising. The level of activity during walking was, in some muscles, lower for the patients with CP than for asymptomatic controls. This is a new finding. CP muscle generated higher mean EMG-frequencies than for asymptomatic controls: this is the opposite of what would be expected if the patients with CP were more fatigued during the assessment. Instead, this result indicates that the muscles in patients with CP have in some way an altered architecture or composition. Imbalances in activity between the tibialis anterior and medial gastrocnemius contributed to drop-foot at the end of swing phase. Patterns of co-activations between antagonistic muscles differed between CP and asymptomatic patients and were EMG-frequency dependent. Muscle dysfunction was greater in the lower than the upper leg. Most importantly, the new analysis method that we developed was able to distinguish the muscle dysfunction in children with spastic CP from asymptomatic controls, with 96 % sensitivity at 95 % specificity.
The new methods thus provide a powerful and sensitive tool for assessing muscle dysfunction with cerebral palsy. The next step will be to investigate whether this tool can be used to predict the success of surgical intervention. If this were the case then it would be of practical use to clinicians for informing about possible treatment options. In order to address this question we have set-up collaborations with the One Small Step Gait Laboratory at Guy’s hospital. At this lab there is EMG data for over 500 patients with CP already assessed. We intend to apply the new methods to this data and use statistical approaches to quantify the power that these methods have for predicting treatment outcomes.
Conclusions
This study has shown that the muscles in children with spastic CP generate EMG signals that differ in timing, magnitude and frequency content from asymptomatic controls.
Wavelet analysis is thus, an effective way to resolve differences in the EMG signals. The differences in EMG signals occur both for individual muscles and in the correlations between antagonistic pairs. We have presented here a methodological framework to quantify these patterns of muscle dysfunction. Using data from this study, these methods have demonstrated that spastic CP muscle generates higher mean EMG frequencies; imbalances in activity between the tibialis anterior and medial gastrocnemius contributed to equinus ankle at the end of swing phase; patterns of coactivations between antagonistic muscles differ between CP and asymptomatic patients and are EMG frequency dependent; muscle dysfunction was greater in the lower than the upper leg; and muscle dysfunction between the tibialis anterior and medial gastrocnemius was distinguished, with 96 % sensitivity at 95 % specificity. We propose that these methods can be used to provide new information about the nature of muscle adaptation to cerebral palsy and to inform good clinical practice for the management of this condition.
This information is not meant to replace the advice of any physician or qualified health professional. The information provided by Cerebra is for information purposes only and is not a substitute for medical advice or treatment for any medical condition. You should promptly seek professional medical assistance if you have concerns regarding any health issue.
Page last updated:
16/12/2011 16:04