Quantitative assessment of upper extremities motor function in multiple sclerosis

Abstract

BACKGROUND:

Upper extremity (UE) motor function deficits are commonly noted in multiple sclerosis (MS) patients and assessing it is challenging because of the lack of consensus regarding its definition. Instrumented biomechanical analysis of upper extremity movements can quantify coordination with different spatiotemporal measures and facilitate disability rating in MS patients.

OBJECTIVE:

To identify objective quantitative parameters for more accurate evaluation of UE disability and relate it to existing clinical scores.

METHODS:

Thirty-four MS patients and 24 healthy controls (CG) performed a finger-to-nose test as fast as possible and, in addition, clinical evaluation kinematic parameters of UE were measured by using inertial sensors.

RESULTS:

Generally, a higher disability score was associated with an increase of several temporal parameters, like slower task performance. The time taken to touch their nose was longer when the task was fulfilled with eyes closed. Time to peak angular velocity significantly changed in MS patients (EDSS $>$ 5.0). The inter-joint coordination significantly decreases in MS patients (EDSS 3.0–5.5). Spatial parameters indicated that maximal ROM changes were in elbow flexion.

CONCLUSIONS:

Our findings have revealed that spatiotemporal parameters are related to the UE motor function and MS disability level. Moreover, they facilitate clinical rating by supporting clinical decisions with quantitative data.

Keywords

Multiple sclerosis finger-to-nose test EDSS quantitative assessment

1. Introduction

Upper extremity (UE) coordination deficits are commonly noted in neurological patients [1, 2, 3, 4]. Specifically, the upper extremity is the basis of our fine motor skills, which are extremely important in the proper execution of activities of daily life [5]. Knowing what parameters of coordination is in healthy subjects UE, it is easy to list the parameters characterizing coordinated motion such as speed, distance, direction, timing, and muscular tension [6].

UE coordination assessment is challenging for clinicians because of the lack of consensus regarding the definition of the coordination assessment procedure. Coordinated movement of the UE is usually described by specific patterns of temporal and spatial variability according to the task. For a UE coordination assessment, there are clinical tests (finger-to-nose (FNT), finger-to-doctor’s finger etc.) which allow the identification of intention tremors and dysmetria [7, 8]. The primary clinical outcome measure for evaluating MS in clinical trials is Kurtzke’s expanded disability status scale (EDSS) [9]. This scale is based on the assessment of eight functional systems and it ranges from 0 to 10 that represent higher levels of disability.

However, the authors suggest that there is a need for additional quantitative methods providing objective and more accurate evaluation of the UE motion in MS patients. In this context, wearable inertial sensors seem to be a good tool for the instrumented analysis of the FNT [10, 11].

We hypothesize that the quantitative evaluation will allow for the severity of the motor disability of the UE in MS patients and we aimed to find a correlation between the UE motor function disability level rated by EDSS and spatiotemporal parameters in MS patients.

2. Methods

2.1 Subjects and clinical assessment

Kinematic data of the UE was collected at the Vilnius University Hospital Santaros Klinikos, Center of Neurology. In total, 54 subjects were enrolled in the study. Twenty-four healthy controls (17 women (71%), mean age $\pm$ standard deviation (SD): 28 $\pm$ 3.97 years and 7 men (29%): 31.14 $\pm$ 5.67 years) and 34 MS patients (21 women (62%): 42.19 $\pm$ 12.55 years and 13 men (38%): 36.46 $\pm$ 13.07 years) who met the following inclusion criteria: (1) multiple sclerosis diagnosis, (2) at least 18 years old, (3) EDSS score, (4) the ability to sit and move the UE, (5) the ability to understand and follow verbal instructions.

Exclusion criteria were: (1) EDSS score $<$ 1.0 points, (2) healthy people with a history of neurological or psychiatric disorders or the use of psychoactive drugs, (3) UE plegia, (4) other diseases or states influencing motor control of upper extremities. The MS group was divided into subgroups MS1, MS2, MS3, MS4, and MS5 with corresponding clinical evaluation according to the EDSS.

Clinical evaluation was based on a neurological examination and the level of disability was measured using the EDSS [9]. EDSS score ranged between 2.0 and 6.5 in all MS patients. The FNT test was performed with eyes open and closed.

All the participants in the study provided their informed consent and the study was approved by the ethic committee of Vilnius University Hospital Santaros Klinikos (protocol No. MS-ATAX-2015-00).

2.2 Experimental setup

The FNT was performed whereas IMUs – wireless 9DOF inertial sensors (Shimmer Research, Dublin, Ireland) were attached to the UE. The kinematic data of the UE was collected. Six sensors were fixed to both UE segments on centres of mass: one on each arm, each forearm, and each hand (Fig. 1). The standardized procedure of FNT was explained and demonstrated by a neurologist for all patients. The subject sat on the chair at starting position (90 ${}^{\circ}$ shoulder flexion, fingers extended, no target) (Fig. 1a).

Table 1
Kinematic measures supplementing the clinical assessment, $n=$ 34

Parameters tested	Kinematic		General FNT test rate
during FNT	parameter	Level of	MS subgroup
		ataxia	MS1	MS2	MS3	MS4	MS5
			$n=$ 11	$n=$ 7	$n=$ 8	$n=$ 5	$n=$ 3
Slowness of movement	$t_{m}$	N/M/S	2/8/1	0/6/1	1/6/1	0/1/4	0/1/2
Slowness of movement to nose touch	$t_{\textit{dif}}$
Clumsy and unsteady movements	$t_{\textit{dif}}$ , $\omega_{\textit{dif}}$ , ROM
Dysmetria of movements, incoordination	$t_{PS}$ , IJC, ROM

Abbreviations: N – no ataxia; M – mild/moderate ataxia; S – severe ataxia.

Figure 1.

Patient setup and task performance: a) position 1; b) position 2.

Table 2

Temporal parameters of MS and CG, $n=$ 58

			Variables
Groups			$t_{\textit{dif}}$ , s	$t_{\textit{PS}}$ , %	$\omega_{\textit{dif}}$ , ${}^{\circ}$ /s
CG $n=$ 24	Open	Right	$-$ 0.08 $\pm$ 0.27 ${}^{\text{c}}$	45.7 $\pm$ 54.6	46.5 $\pm$ 48.3
		Left	0.11 $\pm$ 0.39 ${}^{\text{c}}$	45.7 $\pm$ 48.1	43.9 $\pm$ 35.9
	Closed	Right	0.02 $\pm$ 0.20	42.9 $\pm$ 38.7	40.6 $\pm$ 39.2
		Left	0.00 $\pm$ 0.28	44.0 $\pm$ 68.7	34.1 $\pm$ 33.1
MS1 $n=$ 11	Open	Right	$-$ 0.02 $\pm$ 0.19 ${}^{\text{c}}$	45.1 $\pm$ 33.1	53.0 $\pm$ 47.4
		Left	0.16 $\pm$ 0.15 ${}^{\text{c}}$	55.1 $\pm$ 39.7 ${}^{\text{a}}$	53.8 $\pm$ 47.6
	Closed	Right	0.03 $\pm$ 0.20	46.1 $\pm$ 27.5	31.1 $\pm$ 50.6
		Left	0.17 $\pm$ 0.17 ${}^{\text{a}}$	44.9 $\pm$ 24.8	30.5 $\pm$ 36.1
MS2 $n=$ 7	Open	Right	0.05 $\pm$ 0.19	42.3 $\pm$ 41.4	64.3 $\pm$ 37.1
		Left	0.23 $\pm$ 0.16	66.7 $\pm$ 71.8 ${}^{\text{a}}$	42.2 $\pm$ 45.6
	Closed	Right	0.18 $\pm$ 0.26 ${}^{\text{a}}$	42.5 $\pm$ 34.1	32.2 $\pm$ 28.3
		Left	0.17 $\pm$ 0.24	42.7 $\pm$ 26.0	39.6 $\pm$ 53.1
MS3 $n=$ 8	Open	Right	0.08 $\pm$ 0.24	48.0 $\pm$ 42.7 ${}^{\text{a}}$	26.9 $\pm$ 54.1
		Left	0.25 $\pm$ 0.31	46.3 $\pm$ 26.8	65.3 $\pm$ 34.9
	Closed	Right	0.22 $\pm$ 0.31 ${}^{\text{a}}$	43.2 $\pm$ 46.6 ${}^{\text{a}}$	31.6 $\pm$ 38.7
		Left	0.21 $\pm$ 0.28 ${}^{\text{a}}$	41.1 $\pm$ 27.6	49.9 $\pm$ 18.8
MS4 $n=$ 5	Open	Right	0.11 $\pm$ 0.25 ${}^{\text{a}}$	36.1 $\pm$ 48.6	41.9 $\pm$ 41.1
		Left	0.24 $\pm$ 0.28 ${}^{\text{a}}$	41.2 $\pm$ 35.3	40.8 $\pm$ 26.8
	Closed	Right	0.40 $\pm$ 0.60 ${}^{\text{a}}$	35.1 $\pm$ 24.4 ${}^{\text{ab}}$	41.6 $\pm$ 57.8
		Left	0.45 $\pm$ 0.55 ${}^{\text{a}}$	31.0 $\pm$ 43.1	36.7 $\pm$ 48.6
MS5 $n=$ 3	Open	Right	0.11 $\pm$ 0.10	47.9 $\pm$ 9.40 ${}^{\text{ab}}$	29.0 $\pm$ 17.5
		Left	0.25 $\pm$ 0.08 ${}^{\text{a}}$	53.3 $\pm$ 18.6 ${}^{\text{a}}$	14.1 $\pm$ 42.2
	Closed	Right	0.49 $\pm$ 0.31 ${}^{\text{ab}}$	47.5 $\pm$ 9.70 ${}^{\text{ab}}$	41.2 $\pm$ 12.1
		Left	0.45 $\pm$ 0.11 ${}^{\text{ab}}$	47.7 $\pm$ 8.20 ${}^{\text{ab}}$	23.0 $\pm$ 16.6

All values are represented as mean $\pm$ SD. ${}^{\text{a}}$ – significant difference between CG and MS ( $P<$ 0.05); ${}^{\text{b}}$ – significant difference between MS1 and MS ( $P<$ 0.05); ${}^{\text{c}}$ – significant difference between right and left UE ( $P<$ 0.05).

Figure 2.

Means of movement time ( $t_{m}$ ) in MS (CG shown as base line 2.04 $\pm$ 0.46 (mean $\pm$ SD)), $n=$ 34. ${}^{\text{a}}$ – significant difference between CG and MS ( $P<$ 0.05); ${}^{\text{b}}$ – significant difference between MS1 and MS ( $P<$ 0.05).

Figure 3.

Means of ROMs in MS, $n=$ 58. ${}^{\text{a}}$ – significant difference between CG and MS ( $P<$ 0.05); ${}^{\text{b}}$ – significant difference between MS1 and MS ( $P<$ 0.05).

After an audible signal the subject started to perform the task to touch his/her nose with the tip of index finger as quickly and as accurately as possible (Fig. 1, position 2) and then return the hand to the starting position 1. The task was completed with the right and after with the left arm. All participants repeated the task three times with their eyes open and after three times with closed eyes. The measured data was sampled at a rate of 256 Hz.

2.3 Data analysis

All collected data was analysed using MATLAB software (Mathworks Inc, Natick, USA). The angular velocity from the gyroscope was processed to determine spatiotemporal parameters.

The raw gyroscope signal was filtered using 2nd order Butterworth low-pass filter with a cut-off frequency of 5 Hz and then was normalized.

Temporal parameters were calculated from forearm angular velocity. The time interval between the start of movement to end was defined as movement time ( $t_{m}$ ). Time to peak angular velocity ( $t_{\textit{PS}}$ ) is expressed in percent from start position 1 to $\omega_{\textit{1peak}}$ . Additionally, differences in time and angular velocities were calculated respectively as follows: $t_{\textit{dif}}=t_{\textit{NT}}-t_{R}$ and $\omega_{\textit{dif}}=\omega_{\textit{2peak}}-\omega_{\textit{1peak}}$ .

Spatial parameters were obtained by integrating the filtered raw gyroscope signal. ROM was calculated as difference between the maximum and minimum amplitude of segment.

The following tests were used to verify the existence of a possible relationship between the variables examined: Shapiro-Wilk normality test ( $P<$ 0.05) was used for verifying data normality. Normally distributed data were compared by means of parametric statistical method, i.e. with one-way ANOVA ( $P<$ 0.05). Not normally distributed data ( $P<$ 0.01) were compared by means of non-parametric statistical method, i.e. with the Kruskal-Wallis test ( $P<$ 0.05).

3. Results

In this study, kinematic parameters were correlated only with the clinical assessment by the EDSS score and the FNT clinical test rate. The evaluation of the FNT was supplemented by the kinematic parameters as shown in Table 1.

Analysing obtained temporal parameters (Table 2 and Fig. 2) it was observed that the movement time $t_{m}$ of MS3, MS4, and MS5 (EDSS $>$ 4.0) was significantly longer than in the CG ( $P<$ 0.05) performing FNT with eyes open/closed. However, differences of $t_{m}$ between tasks performed with the open or closed eyes either by left or right UE were not noted in any group. The time difference $t_{\textit{dif}}$ in all MS subgroups differs from CG ( $P<$ 0.05), when the task was fulfilled with eyes closed. The statistically significant difference was evaluated in MS1 and MS5 (eyes closed) and in the MS1 group between right and left UE (eyes open). The $t_{\textit{PS}}$ comparing with the MS1 group there were mostly increases in MS5 (47.5 $\pm$ 9.70% and 47.7 $\pm$ 8.20% both UE/eyes closed, 47.9 $\pm$ 9.40% right UE/eyes open) and decreases in MS4 (35.1 $\pm$ 24.4% right UE/eyes closed).

The $\omega_{\textit{dif}}$ did not display any statistically significant difference neither in MS groups, nor when comparing MS vs CG. However, it was noted that $\omega_{\textit{dif}}$ when eyes were closed, there was a gradual decrease from 42% in MS1 to 5% in MS4 and in MS5 it increased by 2.5%.

Results of joints ROMs (Fig. 3) displayed the following: comparing to the MS1, the elbow flexion in MS2 (both UE/eyes closed), MS3 (both UE), MS4 (eyes closed), and MS5 (both UE/eyes closed) were significantly reduced. In addition, a statistically significant hand rotation (left UE/eyes closed) decrease in the MS4 group and shoulder extension (right UE) in the MS5 group (Fig. 3) were observed.

Compared to CG, a statistically significant increase of ROM was found in the MS1 elbow flexion (eyes opened/closed), and decreasing MS2 shoulder abduction (left UE), MS4 elbow flexion (left UE/eyes closed), as well as in MS5 shoulder extension (right UE/eyes open) (Fig. 3).

Statistically significant differences in ROMs were not observed neither between MS subgroups nor between left/right UE or open/closed eyes.

4. Conclusions

The evaluation of a neurological examination mostly depends on subjective analysis. To evaluate a clinical test such as the FNT is very important to improve the quality of the neurological examination. Despite the limitations, our study showed that measurements with IMUs can greatly facilitate the evaluation of such clinical tests by supporting clinical decisions with quantitative data. Spatiotemporal parameters during the FNT test enable examiners to distinguish MS patients according to EDSS more accurately. In practice, understanding kinematical parameters along with a clinical rating would help to improve the assessment of the degree of MS. Based on these parameters, a real-time system can be developed to help neurologists to objectively evaluate the motor functions of the UE in MS patients.

Footnotes

Conflict of interest

None to report.

References

Rodrigues

MRM

Slimovitch

Chilingaryan

Levin

. Does the Finger-to-Nose Test measure upper limb coordination in chronic stroke? J Neuroeng Rehabil2017; 14: 1–11. doi: 10.1186/s12984-016-0213-y.

Santos

Franco de Moura

Lazzari

Dumont

AJL

Braun

LAF

Oliveira

. Upper limb function evaluation scales for individuals with cerebral palsy: a systematic review. J Phys Ther Sci2015; 27: 1617–20. doi: 10.1589/jpts.27.1617.

Quinn

Busse

Dal Bello-Haas

. Management of upper extremity dysfunction in people with Parkinson disease and Huntington disease: Facilitating outcomes across the disease lifespan. J Hand Ther2013; 26: 148–55. doi: 10.1016/j.jht.2012.11.001.

Delrobaei

Tran

Gilmore

McIsaac

Jog

. Characterization of multi-joint upper limb movements in a single task to assess bradykinesia. J Neurol Sci2016; 368: 337–42. doi: 10.1016/j.jns.2016.07.056.

Ricci

FPFM

Santiago

PRP

Zampar

Pinola

Fonseca M de

. Upper extremity coordination strategies depending on task demand during a basic daily activity. Gait Posture2015; 42: 472–8. doi: 10.1016/j.gaitpost.2015.07.061.

Rab

Petuskey

Bagley

. A method for determination of upper extremity kinematics. Gait Posture2002; 15: 113–9. doi: 10.1016/S0966-6362(01)00155-2.

Alusi

Worthington

Glickman

Findley

Bain

. Evaluation of three different ways of assessing tremor in multiple sclerosis. J Neurol Neurosurg Psychiatry2000; 68: 756–60. doi: 10.1136/jnnp.68.6.756.

Day

Thompson

Harding

Marsden

. Influence of vision on upper limb reaching movements in patients with cerebellar ataxia. Brain1998; 121: 357–72. doi: 10.1093/brain/121.2.357.

Kurtzke

. Rating neurologic impairment in multiple sclerosis. Neurology1983; 33: 1444. doi: 10.1212/wnl.33.11.1444.

10.

Carpinella

Cattaneo

Ferrarin

. Hilbert-Huang transform based instrumental assessment of intention tremor in multiple sclerosis. J Neural Eng2015; 12. doi: 10.1088/1741-2560/12/4/046011.

11.

Shaikh

Zee

Mandir

Lederman

Crawford

. Disorders of Upper Limb Movements in Ataxia-Telangiectasia 2013; 8: 1–7. doi: 10.1371/journal.pone.0067042.