- Email: [email protected]
What's the risk of using the Modified Ashworth Scale (MAS) to assess spasticity at the ankle?
S18
Abstracts / Gait & Posture 33S (2011) S1–S66
Fig. 1. Example of MFPSDF of tibialis anterior SEMG evaluated in 1 CS, 1 T2DM without PN and PAD (D), 1 T2DM with PAD (DV), and 1 T2DM with PAD and PN (DVN).
manifestation of muscle fatigue, determined by means of surface electromyography (SEMG) [2], in dynamic conditions, in a group of type 2 diabetic patients (T2DM). Materials and methods Fourteen T2DM patients and 10 matched control subjects (CS) participated in the study (age 65 ± 10.3; BMI 25.5 ± 3; mean ± SD). Seven T2DM were affected mainly by PAD, 4 by PN and/or AN and 5 were free from complication. Diagnosis of PN, AN and PAD was accurately performed by appropriate evaluations. Patients were then asked to walk on a treadmill for 35 min at a speed of 4 km/h, with an inclination of 2%, after performing 2.5 min of warm-up at 2 km/h [3]. The session ended with a period of cool-down of 2.5 min at 2 km/h. Foot-switch, knee flexion-extension angle and SEMG signals were recorded synchronously by means of STEP32 (DemItalia, Italy). Patients were instrumented bilaterally with foot-switches, knee goniometers and SEMG probes over biceps femoris, vastus lateralis, rectus femoris, tibialis anterior (TA) and gastrocnemius lateralis. SEMG signals were acquired with a sampling frequency of 2 kHz and high-pass filtered at 20 Hz (FIR filter, 100 taps) to attenuate motion artefacts and low-pass filtered at 350 Hz to reduce high-frequency noise. The mean frequency of the Power Spectral Density function (MFPSDF) was estimated on each SEMG at each gait cycle. In order to reduce the estimation variability, the MFPSDF over 30 consecutive gait cycles was averaged. Analysis of correlation was performed (SPSS v13, R software, RODBC package, polychor function) between MFPSDF and clinical parameters in order to highlight possible correlation between muscle fatigue and either NP, AN or PAD. Results Presence of mauscle fatigue was revealed in 8 T2DM (3 with PAD, 1 with PAD and PN, 2 T2DM without any complicances and 2 with NP without PAD) (see Fig. 1). When considering correlation analysis, a nice correlation was found (0.96 > R > 0.6) between the presence of PN, AN, PAD and the presence of electrical muscular fatigue. Discussion Preliminary results show that SEMG can be considered an efficient tool in highlighting the presence of muscle fatigue, during walking in diabetic subjects. The evaluation of the functionality of the lower limb, through the analysis of the manifestation of muscular fatigue, will improve our knowledge of the pathogenesis of PV.
References [1] Van Schie CHM. Int J Low Extrem Wounds 2005;4(3):160–70. [2] McDermott MM, et al. JAMA 2009;301(2):165–74. [3] Chew JT, et al. Clin Orthop Relat Res 1995;320:95–100.
doi:10.1016/j.gaitpost.2010.10.023 O19 What’s the risk of using the Modified Ashworth Scale (MAS) to assess spasticity at the ankle? I. Campanini, A. Merlo, L. Cavazzuti LAM – Laboratorio Analisi Movimento, Dip. Riabilitazione, AUSL Reggio Emilia, Correggio (RE), Italy Introduction Spasticity is one of the major sources of disability in patients with neurological impairment. In the clinical practice and in scientific research either the Ashworth Scale (AS) [1] or its Modified version (MAS) [2] are the most commonly used methods for the measurement of spasticity [3]. However, methodological limitations of this scale are now increasingly being acknowledged [4]. The resistance to passive movement measured by the MAS score is a sum of reflex and active muscle activity and of muscle non-neural passive characteristics. Conversely, a proper assessment of spasticity has to be exclusively associated with an increased stretch reflex activity [4]. Despite being known that the MAS score does not provide a direct measure of muscular reflex activity, it is still used in the decision making process for the treatment of spasticity. A reason for this behaviour could be in the lack, in literature, of a quantitative assessment of the false positive rate of MAS as indicator of spasticity. To clarify this topic, we assessed the false positive rate of the Ashworth Scale when used as a tool for the assessment of spasticity at the ankle. Materials and methods A sample of 80 sub-acute neurological patients (53 M, 27 F, age 56 ± 17 years, average time since lesion 57 ± 21 days) with hemiplegia (39 left, 41 right) were selected according to the following criteria. Inclusion criteria: upper motor neurone syndrome following cerebral vascular accident (CVA), less than 90 days from the vascular accident (sub-acute) to minimise the occurrence of muscle contractures and rigid retractions, unilateral clinical signs. Exclu-
Abstracts / Gait & Posture 33S (2011) S1–S66 Table 1 Comparison between the presence of spasticity at the gastrocnemii (top) and soleus (bottom) muscles as assessed by the presence of reflex EMG activity during the Ashworth manoeuvre and the presence of spasticity as inferred by the presence of a positive MAS score. EMG-assessed spasticity
MAS-based spasticity Yes
No
Total
Yes No
10 23
2 28
12 62
Total
44
30
74
S19
O20 Outcome evaluation of surgery treatment by means of gait analysis in children with cerebral palsy: The normality pattern is not an effective term of comparison R. Neviani 1 , S. Costi 1 , C. Borghi 1 , A. Ferrari 1,2 , S. Faccioli 1 , A. Ferrari 1 1 LAMBDA Motion Laboratory, University of Modena and Reggio Emilia, Reggio Emilia, Italy 2 Xsens Technologies B.V., Enschede, The Netherlands
False detection rate (FDR) 77% EMG-assessed spasticity
MAS-based spasticity Yes
No
Total
Yes No
9 36
2 32
11 68
Total
45
34
79
False detection rate (FDR) 80%
sion criteria: concomitant medullar lesion, akinesia or iperkinesia, hydrocephalus, cerebellar syndrome. Patients with elicitable ankle clonus were also excluded. A single experienced rater assessed the MAS score for all patients, at their bed, for gastrocnemius medialis (GM) and lateralis (GL), and soleus (SOL) muscles. The presence of reflex muscular activity was measured by surface EMG, used as the criterion standard to assess the presence of spasticity. Bi-polar surface electrodes (ARBO H124 SG, diameter 15 mm, interelectrode distance 20 mm) were placed on the minimum crosstalk points recommended in [5]. EMG data were collected at 1000 Hz (PocketEMG, BTS, Italy). During the MAS assessment sEMG data were acquired but not displayed to ensure the assessor’s blindness to the presence of EMG. The presence in sEMG traces of muscular activity consequent to the stretching manoeuvres was assessed separately by visual inspection by an experienced operator. This assessor was blind to any information on patients and, mainly, to all MAS scores. A score “Yes” was given to trials with EMG reflex activity, and a score “No” was given otherwise. Results Results are presented in Table 1. Discussion We investigated the diagnostic risk in the use of the Ashworth Scale as a tool for inferring the presence of spasticity at the ankle in a sample of 80 adult CVA patients. The global rate of false positive, that is the percentage of positive MAS scores assigned to patients who did not have a reflex activity at the MAS manoeuvre, approached 80% (Table 1). In practice, if one would suppose the presence of spasticity at the calf based on a positive MAS score, an error will be made on three CVA patients out of four. MAS does not reflect the presence of spasticity at the ankle in adults who have had a stroke. The treatment of spasticity at the ankle cannot rely on the Ashworth score for adult CVA patients. References [1] [2] [3] [4] [5]
Ashworth B. Practitioner 1964;192:540–2. Bohannon RW, et al. Phys Ther 1987;67(2):206–7. Platz, et al. Disabil Rehabil 2005;27:7–18. Alibiglou, et al. J Neuroeng Rehabil 2008;5:18. Basmajian JV, DeLuca CJ. Muscles alive. Baltimore: Williams & Wilkins; 1985.
doi:10.1016/j.gaitpost.2010.10.024
Introduction Among the treatments available for the rehabilitation of the alterations of walking in children affected by cerebral palsy (CP), multilevel surgery is one of the most effective and used solutions despite its irreversible nature. Literature provides increasingly evidences of the large benefits afforded by gait analysis (GA) in the process of surgery decision making and in the choice of the most appropriate surgery tecnique [1]. However, besides the diagnostic process, GA can be used in order to measure accurately and effectively the outcome of the treatment [1]. The aim of the present study is to demonstrate how deviations in the kinematic pattern of walking provoked by a surgery treatment can determine (i) an improvement in the motor performance and (ii) an increase in the autonomy level in action, also when they vary in the opposite direction with respect to the normal pattern. Materials and methods 15 dipelgic children with purely spastic forms of CP (age range 7–17 years) classified in one of the four forms of diplegia proposed by [2] and addressed by a clinical examination to undergo a functional surgery treatment on the lower limbs, participated in the study. The aim of the surgery treatment was to improve the motor performance and the autonomy level in walking. The treatments adopted included tendon and muscular lengthenings, and interventions of skeleton correction. The motor performance exhibited by the subjects was measured both by means of the Gross Motor Function Measure 88 (GMFM) by using only the dimension D and E [3], and of GA, the day before (session PRE) and after six months (session POST) the surgery intervention. The kinematics of at least three gait cycles for each limb was acquired through the protocol Total3DGait [4] by means of an 8 cameras Vicon MX+ system (Vicon Motion System, UK). The 9 kinematic variables relative to the sagittal, frontal and transverse plane rotations of the hip, knee and ankle and the 4 kinematic variables relative to the rotations on the sagittal and frontal plane of the segments pelvis and trunk obtained during the PRE and POST sessions were compared with respect to the normality bands [4]. In particular, for each of the 13 kinematic variables considered, sets of 3 mean waveforms were computed from the gait cycles acquired in the two sessions PRE and POST and from the normality bands. For each subject, the approach of the 13 mean waveforms of the POST session towards the normality, with respect to the ones of the session PRE, was computed in terms of offset (Off) and Range of Movement (ROM). In particular, positive values of Off and ROM indicate that the POST mean waveforms move closer to the normal ones in both absolute and range terms. Results The obtained GMFM values were positive in 13 out of 15 cases, with a percentage mean value of 3.4 and a standard deviation of 6.1. Table 1 reports the mean and the standard deviation values relative to Off and ROM obtained from the 13 kinematics variable considered for each of the 15 enrolled subjects. Discussion The obtained GMFM data, revealed an improvement in functional autonomy level in action in 13 out of 15 cases. On the
Abstracts / Gait & Posture 33S (2011) S1–S66
Fig. 1. Example of MFPSDF of tibialis anterior SEMG evaluated in 1 CS, 1 T2DM without PN and PAD (D), 1 T2DM with PAD (DV), and 1 T2DM with PAD and PN (DVN).
manifestation of muscle fatigue, determined by means of surface electromyography (SEMG) [2], in dynamic conditions, in a group of type 2 diabetic patients (T2DM). Materials and methods Fourteen T2DM patients and 10 matched control subjects (CS) participated in the study (age 65 ± 10.3; BMI 25.5 ± 3; mean ± SD). Seven T2DM were affected mainly by PAD, 4 by PN and/or AN and 5 were free from complication. Diagnosis of PN, AN and PAD was accurately performed by appropriate evaluations. Patients were then asked to walk on a treadmill for 35 min at a speed of 4 km/h, with an inclination of 2%, after performing 2.5 min of warm-up at 2 km/h [3]. The session ended with a period of cool-down of 2.5 min at 2 km/h. Foot-switch, knee flexion-extension angle and SEMG signals were recorded synchronously by means of STEP32 (DemItalia, Italy). Patients were instrumented bilaterally with foot-switches, knee goniometers and SEMG probes over biceps femoris, vastus lateralis, rectus femoris, tibialis anterior (TA) and gastrocnemius lateralis. SEMG signals were acquired with a sampling frequency of 2 kHz and high-pass filtered at 20 Hz (FIR filter, 100 taps) to attenuate motion artefacts and low-pass filtered at 350 Hz to reduce high-frequency noise. The mean frequency of the Power Spectral Density function (MFPSDF) was estimated on each SEMG at each gait cycle. In order to reduce the estimation variability, the MFPSDF over 30 consecutive gait cycles was averaged. Analysis of correlation was performed (SPSS v13, R software, RODBC package, polychor function) between MFPSDF and clinical parameters in order to highlight possible correlation between muscle fatigue and either NP, AN or PAD. Results Presence of mauscle fatigue was revealed in 8 T2DM (3 with PAD, 1 with PAD and PN, 2 T2DM without any complicances and 2 with NP without PAD) (see Fig. 1). When considering correlation analysis, a nice correlation was found (0.96 > R > 0.6) between the presence of PN, AN, PAD and the presence of electrical muscular fatigue. Discussion Preliminary results show that SEMG can be considered an efficient tool in highlighting the presence of muscle fatigue, during walking in diabetic subjects. The evaluation of the functionality of the lower limb, through the analysis of the manifestation of muscular fatigue, will improve our knowledge of the pathogenesis of PV.
References [1] Van Schie CHM. Int J Low Extrem Wounds 2005;4(3):160–70. [2] McDermott MM, et al. JAMA 2009;301(2):165–74. [3] Chew JT, et al. Clin Orthop Relat Res 1995;320:95–100.
doi:10.1016/j.gaitpost.2010.10.023 O19 What’s the risk of using the Modified Ashworth Scale (MAS) to assess spasticity at the ankle? I. Campanini, A. Merlo, L. Cavazzuti LAM – Laboratorio Analisi Movimento, Dip. Riabilitazione, AUSL Reggio Emilia, Correggio (RE), Italy Introduction Spasticity is one of the major sources of disability in patients with neurological impairment. In the clinical practice and in scientific research either the Ashworth Scale (AS) [1] or its Modified version (MAS) [2] are the most commonly used methods for the measurement of spasticity [3]. However, methodological limitations of this scale are now increasingly being acknowledged [4]. The resistance to passive movement measured by the MAS score is a sum of reflex and active muscle activity and of muscle non-neural passive characteristics. Conversely, a proper assessment of spasticity has to be exclusively associated with an increased stretch reflex activity [4]. Despite being known that the MAS score does not provide a direct measure of muscular reflex activity, it is still used in the decision making process for the treatment of spasticity. A reason for this behaviour could be in the lack, in literature, of a quantitative assessment of the false positive rate of MAS as indicator of spasticity. To clarify this topic, we assessed the false positive rate of the Ashworth Scale when used as a tool for the assessment of spasticity at the ankle. Materials and methods A sample of 80 sub-acute neurological patients (53 M, 27 F, age 56 ± 17 years, average time since lesion 57 ± 21 days) with hemiplegia (39 left, 41 right) were selected according to the following criteria. Inclusion criteria: upper motor neurone syndrome following cerebral vascular accident (CVA), less than 90 days from the vascular accident (sub-acute) to minimise the occurrence of muscle contractures and rigid retractions, unilateral clinical signs. Exclu-
Abstracts / Gait & Posture 33S (2011) S1–S66 Table 1 Comparison between the presence of spasticity at the gastrocnemii (top) and soleus (bottom) muscles as assessed by the presence of reflex EMG activity during the Ashworth manoeuvre and the presence of spasticity as inferred by the presence of a positive MAS score. EMG-assessed spasticity
MAS-based spasticity Yes
No
Total
Yes No
10 23
2 28
12 62
Total
44
30
74
S19
O20 Outcome evaluation of surgery treatment by means of gait analysis in children with cerebral palsy: The normality pattern is not an effective term of comparison R. Neviani 1 , S. Costi 1 , C. Borghi 1 , A. Ferrari 1,2 , S. Faccioli 1 , A. Ferrari 1 1 LAMBDA Motion Laboratory, University of Modena and Reggio Emilia, Reggio Emilia, Italy 2 Xsens Technologies B.V., Enschede, The Netherlands
False detection rate (FDR) 77% EMG-assessed spasticity
MAS-based spasticity Yes
No
Total
Yes No
9 36
2 32
11 68
Total
45
34
79
False detection rate (FDR) 80%
sion criteria: concomitant medullar lesion, akinesia or iperkinesia, hydrocephalus, cerebellar syndrome. Patients with elicitable ankle clonus were also excluded. A single experienced rater assessed the MAS score for all patients, at their bed, for gastrocnemius medialis (GM) and lateralis (GL), and soleus (SOL) muscles. The presence of reflex muscular activity was measured by surface EMG, used as the criterion standard to assess the presence of spasticity. Bi-polar surface electrodes (ARBO H124 SG, diameter 15 mm, interelectrode distance 20 mm) were placed on the minimum crosstalk points recommended in [5]. EMG data were collected at 1000 Hz (PocketEMG, BTS, Italy). During the MAS assessment sEMG data were acquired but not displayed to ensure the assessor’s blindness to the presence of EMG. The presence in sEMG traces of muscular activity consequent to the stretching manoeuvres was assessed separately by visual inspection by an experienced operator. This assessor was blind to any information on patients and, mainly, to all MAS scores. A score “Yes” was given to trials with EMG reflex activity, and a score “No” was given otherwise. Results Results are presented in Table 1. Discussion We investigated the diagnostic risk in the use of the Ashworth Scale as a tool for inferring the presence of spasticity at the ankle in a sample of 80 adult CVA patients. The global rate of false positive, that is the percentage of positive MAS scores assigned to patients who did not have a reflex activity at the MAS manoeuvre, approached 80% (Table 1). In practice, if one would suppose the presence of spasticity at the calf based on a positive MAS score, an error will be made on three CVA patients out of four. MAS does not reflect the presence of spasticity at the ankle in adults who have had a stroke. The treatment of spasticity at the ankle cannot rely on the Ashworth score for adult CVA patients. References [1] [2] [3] [4] [5]
Ashworth B. Practitioner 1964;192:540–2. Bohannon RW, et al. Phys Ther 1987;67(2):206–7. Platz, et al. Disabil Rehabil 2005;27:7–18. Alibiglou, et al. J Neuroeng Rehabil 2008;5:18. Basmajian JV, DeLuca CJ. Muscles alive. Baltimore: Williams & Wilkins; 1985.
doi:10.1016/j.gaitpost.2010.10.024
Introduction Among the treatments available for the rehabilitation of the alterations of walking in children affected by cerebral palsy (CP), multilevel surgery is one of the most effective and used solutions despite its irreversible nature. Literature provides increasingly evidences of the large benefits afforded by gait analysis (GA) in the process of surgery decision making and in the choice of the most appropriate surgery tecnique [1]. However, besides the diagnostic process, GA can be used in order to measure accurately and effectively the outcome of the treatment [1]. The aim of the present study is to demonstrate how deviations in the kinematic pattern of walking provoked by a surgery treatment can determine (i) an improvement in the motor performance and (ii) an increase in the autonomy level in action, also when they vary in the opposite direction with respect to the normal pattern. Materials and methods 15 dipelgic children with purely spastic forms of CP (age range 7–17 years) classified in one of the four forms of diplegia proposed by [2] and addressed by a clinical examination to undergo a functional surgery treatment on the lower limbs, participated in the study. The aim of the surgery treatment was to improve the motor performance and the autonomy level in walking. The treatments adopted included tendon and muscular lengthenings, and interventions of skeleton correction. The motor performance exhibited by the subjects was measured both by means of the Gross Motor Function Measure 88 (GMFM) by using only the dimension D and E [3], and of GA, the day before (session PRE) and after six months (session POST) the surgery intervention. The kinematics of at least three gait cycles for each limb was acquired through the protocol Total3DGait [4] by means of an 8 cameras Vicon MX+ system (Vicon Motion System, UK). The 9 kinematic variables relative to the sagittal, frontal and transverse plane rotations of the hip, knee and ankle and the 4 kinematic variables relative to the rotations on the sagittal and frontal plane of the segments pelvis and trunk obtained during the PRE and POST sessions were compared with respect to the normality bands [4]. In particular, for each of the 13 kinematic variables considered, sets of 3 mean waveforms were computed from the gait cycles acquired in the two sessions PRE and POST and from the normality bands. For each subject, the approach of the 13 mean waveforms of the POST session towards the normality, with respect to the ones of the session PRE, was computed in terms of offset (Off) and Range of Movement (ROM). In particular, positive values of Off and ROM indicate that the POST mean waveforms move closer to the normal ones in both absolute and range terms. Results The obtained GMFM values were positive in 13 out of 15 cases, with a percentage mean value of 3.4 and a standard deviation of 6.1. Table 1 reports the mean and the standard deviation values relative to Off and ROM obtained from the 13 kinematics variable considered for each of the 15 enrolled subjects. Discussion The obtained GMFM data, revealed an improvement in functional autonomy level in action in 13 out of 15 cases. On the