Hand Functional Recovery in Sub-acute Brain Injury Stage Patients

using AMADEO® Robotic-assisted Therapy

A Pilot Clinical Study with Apraxic and Neglect Patients

Pedro A. Serrano López-Terradas,

Diana Moya Rosendo

and Marcos Ríos Lago

1,2,3

Unidad de Daño Cerebral, Hospital Beata María Ana de Jesús, Madrid, Spain

Dpto. Psicología Básica II, UNED, Madrid, Spain

Fundación Reina Sofía-Fundación CIEN, Madrid, Spain

Keywords: Amadeo Robot, Brain Injury, Hand Functional Rehabilitation, Robot-assisted Therapy, Stroke.

Abstract: INTRODUCTION: Repeated and intensive exercise with AMADEO® Robot-Assisted Therapies (RAT) has

been found useful in restoring functions of hand paresis in some brain-injured patients. OBJECTIVES:

Evaluate the effects of RAT using AMADEO® device in combination with Conventional Neuro-

rehabilitation Therapy (CNT) for the hand functional recovery in post-acute phase patients and to identify

differences in the hand outcome trends among infants and adults with different recovery potential.

METHODS: 12 adults and 3 infants with neglect or apraxia and hemi-paresis of the upper limb were

enrolled in this prospective randomized pre-post pilot clinical study. They were assigned a priori to positive

(PF+) or negative (PF-) prognostic factor groups. All subjects followed the same standardized protocol with

AMADEO® and CNT. The outcome measures selected were: ARAT, MAS, COTNAB (subtest III), RASP,

RPAB, and AMADEO® ROM and Strength Assessment Tools. RESULTS: Statistical analysis showed

important differences between PF+ and PF- groups in hand function outcome measurements. Similar

improved trends were found between PF+ and the group of infants. Both groups improved in extension

variables, total score, level of difficulty, and speed in performing robot-graded tasks. They also showed

more strength and motor control. Patients in the PF- group showed only hand recovery in flexion and ROM

variables after using the robotic device. Positive intra- and intersession effects were found in all patients.

DISCUSSION: The results suggest that finger motor activation and less somato-sensorial impairments in

pre-test could be a better sign for the prognosis of hand recovery and for the decision to apply Amadeo® in

opposite to the presence or absence of apraxic or neglect symptoms, which have been referred as

contraindications. Amadeo® was a valuable tool, easy to use, safe and useful to monitor hand recovery and

improve grip and finger motor function in spite of the presence of other cognitive impairments.

1 INTRODUCTION

Nowadays, improving hand hemiparesis after brain

injury is a main objective in neuro-rehabilitation in

order to decrease disability in post- brain-injured

survivors. Chronic hand paresis deficits are

prevalent in the distal upper extremities in over 40 %

of individuals, especially regarding arm and hand

motor function (Wang, 2012). Most studies have

found that proximal improvements do not migrate to

the distal arm or vice versa (Takahashi, 2008).

Unfortunately, some of these patients with potential

for partial hand recovery could be excluded from

using robotic devices such as AMADEO® due to

interference with other cognitive impairments, as

apraxia or neglect. These symptoms have been

referred in the instruction Tyromotion manual of

AMADEO® as contraindications. Repeated and

intensive exercise with robot-assisted therapies has

been found useful in restoring functions of upper

extremities by their ability to deliver well-defined

repetitive exercises consistently. Furthermore, the

highly frequent afferent stimulation combined with

increments in efferent activity can stimulate the

mirror neurons and can lead to a shift in the

contribution of the Sensory-Motor Cortex (SMC)

activation of the unlesioned and lesioned

hemispheres (Enzinger, 2012).

Serrano López-Terradas P., Moya Rosendo D. and Ríos Lago M..

Hand Functional Recovery in Sub-acute Brain Injury Stage Patients using AMADEO

 Robotic-assisted Therapy - A Pilot Clinical Study with Apraxic

and Neglect Patients.

 2013 SCITEPRESS (Science and Technology Publications, Lda.)

2 OBJECTIVES

The first aim was to evaluate the effects of robotic-

assisted therapy (RAT) using AMADEO® hand

device in combination with occupational and

physiotherapy conventional neuro-rehabilitation

(CNT) for the hand functional recovery in post-acute

phase (Figure 1).

The second aim was to identify differences in the

hand outcome trends among infants and adults with

different recovery potential. Finally, safety

contraindications of the standardized guidelines of

the AMADEO® Tyromotion manual, version R5

(2010 - 2011), will be discussed.

Figure 1: Use of the AMADEO® hand robot device.

3 METHODS

A sample of 12 adults and 3 infants with moderate to

high grade of hemi-paresis of the upper limb, an

Asworth Scale score of spasticity minor than 3 and

an evolution time inferior to 6 months (Table 1)

were enrolled in this prospective randomized pre-

post pilot clinical study.

On one side, all patients with any extension or

flexion hand activity and less somato-sensorial

disability (moderate impairment in the upper limb)

were assigned a priori to group 1, with a positive

prognostic factor (PF+). On the other side, patients

without voluntary finger activation and high somato-

sensorial disability (severe impairment in the upper

limb) were assigned to group 2, with a negative

prognosis factor (PF-). Apraxic or neglect

impairments were not kept in mind for patient

grouping. The 3 children were included on a third

independent group.

All subjects followed the same standardized

protocol (45 min per session, 2-3 times a week, 12

weeks maximum) with the hand robot, i.e., passive

(CPM), active-assisted (AAT) or active task-

oriented (AT) repetitive hand/finger trainings.

Throughout the 3 months of treatment, all patients

received similar conventional multidisciplinary

neuro-rehabilitation and specific occupational and

physiotherapy sessions to optimize hand functional

sensory-motor performance.

Table 1: Demographic variables of sample groups. CVA:

cerebrovascular accident; Infants: Infants group; L: left;

PF+: Positive prognosis factor group; PF-: Negative

prognosis factor group; R: right; TBI: traumatic brain

injury.

PF+ PF- Infants

Age (years) 63 ± 11 56 ± 12 12 ± 3

Etiology

1 TBI

5 CVA

6 CVA 3TBI

Hemi-paresis

1 R

5 L

3 R

3 L

1 R

2 L

Hand dominance

5 R

1 L

5 R

1 L

3 R

0 L

The following primary and secondary hand

function outcome measures (Sivan, 2011) were

selected: Action Research Arm Test (ARAT) to

assess activities, and Motor Assessment Scale

(MAS), along with Chessington Occupational

Therapy Sensory-motor Assessment (COTNAB-

subtest III) and Rivermead Assessment of

Somatosensory Performance (RASP), to assess body

function. AMADEO® Range of Motion (ROM) and

Strength Assessment (SA) tools were also used as

pre-post outcome measures. Rivermead Perceptual

Assessment Battery (RPAB) and Rivermead

Assessment of Somatosensory Performance (RASP)

were used prior to the use of this robotic device to

diagnose apraxic or neglect symptoms in all patients.

4 RESULTS

Assuming normality, homocedasticity and sphericity

in all variables, preliminary data analysis by

ANOVA of repeated measurements followed by

post-hoc tests showed not statistically significant

differences between PF+ and the group of infants in

any of the variables analysed. Moreover, similar

improved trends of recovery were found in these two

groups.

In most of the variables analyzed, statistical

significant differences were found for the effect of

treatment between PF+ and PF- groups (p < 0.05).

However, the statistical analysis of the

interaction among groups showed, with a high

contrast potency, that only PF+ and the group of

infants improved in the extension variables such as

total extension SA score, individual II-III-IV-V

finger activation in extension SA and extension

trend of recovery (p < 0.05).

With a moderate contrast potency, there were no

differences among both groups in quantity of

improvements associated to flexion variables (p >

0.05) such as flexion SA total score, flexion SA pre-

test, individual I-II-III-IV-V finger activation in

flexion SA, and flexion trend of recovery. Also, the

effect of the interaction within and between subjects

was significant in those variables.

Further post hoc analyses showed a more

significant improvement in the PF+ group than PF-

group in pre-post measurements. The PF- group also

showed hand recovery although less significantly in

variables such as flexion SA total score, ROM and

flexion recovery trends after using robotic-device

(Figure 2). Additionally, in 3 out of 6 PF- cases we

found progressively lower hand tone intra- and inter-

sessions. Besides, the PF+ group improved in the

score, level of difficulty and speed in performing

robot-graded tasks (Figure 3). It also showed more

strength and motor control (Norouzi-Gheidari,

2012).

Positive intra- and inter-session effects were

found in all patients, particularly good tolerance,

motivation and absence of pain.

Patients in the PF+ group with contraindications

to apply AMADEO® (e.g., apraxia and symptoms of

neglect) improved hand motor function and

increased the use of affected hand in the post-tests.

In contrast, patients with lowest somato-sensory

performance showed worse hand outcomes. Some

patients with PF- a priori and apraxia or neglect

symptoms demonstrated smaller but positive

outcomes.

Figure 2: Hand function assessment with AMADEO®.

Em: mean Extension; Fm: mean Flexion; PF+: Positive

prognosis group; PF-: Negative prognosis group.

Important differences between both groups in

hand function outcome measurements were found

for COTNAB-III (p = 0.000), MAS (p = 0.000) and

ARAT (p = 0,000) tools at post-tests (Figure 4).

Figure 3: Pre-post hand function improvements with

AMADEO® tool. ROM: Range of motion. *Statistically

significant differences (p < 0.05).

Figure 4: Main hand outcome measurements. Please, see

Methods for abbreviations on the variables used. Pre: pre-

test; Pos: post-test. *Statistically significant differences (p

< 0.05).

5 DISCUSSION

The findings of this pilot study seem to reveal that

the safety contraindications to apply clinically this

robot device (page 6, version R5, 2010 - 2011)

should be revised. The results suggest that minimal

finger activation motor skills and less somato-

sensorial impairments, such as baseline in pre-test,

could be a better sign for the prognosis of hand

recovery and the decision to apply Amadeo® in

opposite to the presence or absence of apraxia or

neglect symptoms. For this group of patients with

cognitive impairments, Amadeo® was a valuable

tool with a complex intervention design (Hwang,

2012), easy to use, safe and useful to monitor hand

recovery and improve hand grip and finger motor

function.

Our pilot results evidence a better recovery

prognosis for children or patients with motor finger

activation in the hand and less somatosensorial

deficiency, contrary to patients with severe sensitive

damage and serious hand paresis. Apraxic and

neglect symptoms can interfere and complicate the

recovery of the paretic hand, but it is not decisive.

Patients with neglect could benefit from the AT

program facilitated by the adjustments of the

AMADEO® software, whereas patients with apraxia

could improve their motor control due to the

combination of AAT and AT programs.

The combined use of CNT, RAT and splinting

has allowed to regulate and reduce the flexor tone of

hand in 3/12 adult cases.

The presence of other concomitant variables and

their potential positive or negative contribution to

hand function recovery, such as cranioplasty, cancer

radiotherapy or pneumonia due to dysphagia, has

been observed although not analyzed during this

pilot study.

Despite lacking finger opposition and ADD-

ABD of thumb and lumbrical and interosea

movements, the appropriate visual, auditory and

somatosensorial feedback of motor execution, along

with the possibility of working the finger

discrimination of movements and the feedforward at

the higher levels of the device, transforms

AMADEO in a good tool to improve the hand

function in combination with CNT.

The total or partial improvements of one or

several study groups (PF+, PF- and infants) in the

main variables of motor function analyzed (ROM,

flexion or extension SA, finger activation, activities,

and body functions) justifies the necessity of new

studies. More studies will be needed to assess impact

of our results on the activities of daily living (ADL).

Specifically, the small sample size and the absence

of a control group in this study did not allow us to

verify whether this treatment is valid in terms of

effectiveness and universality (Sale, 2012). In future

studies, more statistical analysis will be needed to:

(1) further estimate the impact of these results on

larger samples, (2) to compare with a control group

the outcomes between RAT and CNT with a broader

variety of time and intensity regimens (MacClelland,

2012), and (3) to assess the generalization of

outcomes on ADL with a repetitive, functional and

specific task-oriented rehabilitation in order to

ensure hand function improvements in brain-injured

patients. The good outcomes found in this pilot

clinical study encourage us to design a larger

prospective study.

REFERENCES

Enzinger, C., Pargfrieder, C. Pegritz, S., Wurm, W.,

Linderl-Madrutter, R., Reiter, G., et al., 2012. A proof

of concept study on the effects of a robotic-assisted

hand rehabilitation programme after stroke oncentral

movement control. Conference Meeting.

Hwang, C. H., Seong, J. W., Son, D. S., 2012. Individual

Finger synchronized robot-assisted hand rehabilitation

in subacute to chronic stroke: a prospective

randomized clinical trial of efficacy. Clinical

Rehabilitation 0, 1.

MacClelland, L., Bradham, D., Whitall, J., Volpe, B.,

Wilson, P. D., Ohlhoff, J. et al., 2012. Robotic upper-

limb rehabilitation in chronic stroke patients. J.

Rehabil. Res. Develop. 42, 6, 717.

Norouzi-Gheidari, N., Archambault, P., Fung, J., 2012.

Effects of robot-assisted therapy on stroke

rehabilitation in upper limbs: Systematic review and

meta-analysis of the literature. J. Rehabil. Res.

Develop. 49, 4, 479.

Sale, P., Lombardi, V., Franceschini, M, 2012. Hand

Robotics Rehabilitation: Feasibility and Preliminary

results of a robotic Treatment in patients with

hemiparesis. Stoke Res. Treatment article ID 820931.

Sivan, M., O´Connor, R. J., Makower, S., Levesley, M.,

Bhakta, B., 2011. Systematic review of outcome

measures used in the evaluation of robot-assitested

upper limb exercise in stroke. J. Rehab. Med. 43, 181.

Takahashi, C. D., Der-Yeghiaian, L., Le, V., Motiwala, R.

R., Cramer, S. C., 2008. Robot-based hand motor

therapy after stroke. Brain 131, 425.

Wang, F., 2012. Recent research advances in upper-

extremity rehabilitation. J. Adv. Robot. Automat. 1, 3.