Hand rehabilitation with sonification techniques in the subacute stage of stroke

ABSTRACT After a stroke event, most survivors suffer from arm paresis, poor motor control and other disabilities that make activities of daily living difficult, severely affecting quality of

life and personal independence. This randomized controlled trial aimed at evaluating the efficacy of a music-based sonification approach on upper limbs motor functions, quality of life and

pain perceived during rehabilitation. The study involved 65 subacute stroke individuals during inpatient rehabilitation allocated into 2 groups which underwent usual care dayweek)

respectively of standard upper extremity motor rehabilitation or upper extremity treatment with sonification techniques. The Fugl-Meyer Upper Extremity Scale, Box and Block Test and the

Modified Ashworth Scale were used to perform motor assessment and the McGill Quality of Life-it and the Numerical Pain Rating Scale to assess quality of life and pain. The assessment was

performed at baseline, after 2 weeks, at the end of treatment and at follow-up (1 month after the end of treatment). Total scores of the Fugl-Meyer Upper Extremity Scale (primary outcome

measure) and hand and wrist sub scores, manual dexterity scores of the affected and unaffected limb in the Box and Block Test, pain scores of the Numerical Pain Rating Scale (secondary

outcomes measures) significantly improved in the sonification group compared to the standard of care group (time*group interaction < 0.05). Our findings suggest that music-based

sonification sessions can be considered an effective standardized intervention for the upper limb in subacute stroke rehabilitation. SIMILAR CONTENT BEING VIEWED BY OTHERS A RANDOMISED

CLINICAL TRIAL COMPARING 35 HZ VERSUS 50 HZ FREQUENCY STIMULATION EFFECTS ON HAND MOTOR RECOVERY IN OLDER ADULTS AFTER STROKE Article Open access 28 April 2021 FEASIBILITY OF HOME HAND

REHABILITATION USING MUSICGLOVE AFTER CHRONIC SPINAL CORD INJURY Article Open access 09 November 2022 A SOFT SUPERNUMERARY HAND FOR REHABILITATION IN SUB-ACUTE STROKE: A PILOT STUDY Article

Open access 13 December 2022 INTRODUCTION After a stroke event, most survivors (80–90%) suffer from arm paresis, poor motor control and other disabilities, which evolve in a chronic

condition in about 30–40% of cases1,2. Consequently, patients have difficulty in performing activities of daily living, a fact that severely affects their quality of life and independence3.

Task-oriented sensory-motor training, which allows to transmit the sensory information of the feedback to the central nervous system during task execution, and movements adaptation, is

recognized as significant in increasing post-stroke arm function and dexterity2,4. Beside other factors, such as patient’s compliance and subjective interest in stimulus, the intensity of

training and patient’s motivation have been indicated as key features for a successful rehabilitation therapy5,6. The development of technologies for rehabilitation has made it possible to

formulate new application paradigms obtained by integrating the current rehabilitation pathways with instrumental interventions7,8. Sensory-motor rehabilitation techniques, obtained through

technological devices (such as virtual reality, robots, non-invasive stimulations , motion capture) and used in support of traditional rehabilitation techniques, seem to provide objective

parameters for patient evaluation, accelerate the process of motor recovery and improve motor performance at discharge by means of a top-down approach9,10. Although the results obtained with

the currently available devices are encouraging, we are only at an early stage for the exploitation of these technologies. In fact, while technology-assisted rehabilitation of the upper

limb has demonstrated to have a significant impact on motor outcomes, especially at the proximal level (shoulder and elbow)11, the functional recovery of the hand (in particular of the

prehension function) still presents some issues12,13. Music and its elements (especially the rhythm) are widely used in rehabilitation14.This is due to the activation of the neuronal

networks, in the limbic and paralimbic areas, and in the brain areas involved in movement (motor cortex, supplementary motor area, cerebellum, basal ganglia, etc.)15,16. Many studies

document the possibility that exposure to music during training, but, also through specific rehabilitation interventions, may induce plastic changes15,16,17,18,19 in the sensory-auditory

circuits and motor areas20,21. These changes, resulting in a neuronal reorganization in the nodal points of the brain networks and in fiber bundles, can determine effects lasting beyond the

actual duration of the rehabilitation intervention15. Moreover, music in the rehabilitation process determines an emotional involvement, creating a strong motivational basis15,16. The recent

literature shows how the use of music in stroke rehabilitation can improve gait (speed, cadence and step length, balance)22,23, movement of the upper limbs24,25, language26,27, but also

mood and other psychological aspects28,29,30. More recent studies use, for the rehabilitation of the upper limbs, the mapping of the patient’s movements to whom a sonorous stimulus is

associated (sonification)31,32,33,34. Sonification can improve motor functions rehabilitation and can facilitate the integration of auditory and sensory-motor systems35,36,37. This technique

can also strengthen and support the damaged proprioceptive system and can make the rehabilitative process more pleasant and stimulating from an emotional and motivational point of

view15,37. In some studies, the kinematic data of the gesture are translated into modulations of some parameters (typically frequency and amplitude) of a continuous synthetic sound38,39,40.

In other post-stroke hand rehabilitation studies, the sonification action was based on the execution of short scales or melodic intervals that the patient had to reproduce through specific

movements, with the final goal of creating "simple nursery rhymes or other familiar tunes", modulating music timbre and intensity41. Further, Reh et al.42 supported the

rehabilitation of walking by a combination of an ecological sound (noise produced by a step in the snow) together with a predetermined glissando effect. In this study the sonification can be

considered as a properly selected set of sonorous-music stimuli activated by patient’s movements with the mediation of a sensor (in this specific case the Leap Motion Controller).

Synthesized sounds/musical texture and their parameters (mainly rhythm, pitch/melody, intensity/dynamics, harmony and timbre) are used to represent movements characteristics, especially from

a temporal and spatial point of view. Sonification makes possible the improvement of sensorimotor learning, proprioception, movements planning and execution37. In particular music elements

used in this study were automatically associated to the patient’s movements without involvement of any cognitive tasks. This allows the patient to focus attention mainly on the motor

outcome. In particular, given the importance of hand functions in activities of daily living and manipulation, the protocol (Sonichand) was developed and validated for the training of

pronation and supination of the forearm, ulnar and radial deviation of the wrist and hand grasping movements43. The main objective of this randomized controlled study conducted in subacute

stroke patients during inpatient rehabilitation, was to verify the effectiveness (evaluated through the assessment of the upper limb level of impairment at baseline and end of treatment) of

a new rehabilitative hand treatment based on a musical sonification approach compared to a conventional intervention. The secondary aim was to evaluate if this technique could be beneficial

for reduction of pain perceived during training and improve the patient's perceived quality of life. The main hypothesis of this study was that our sonification approach could be more

effective than a conventional intervention for the rehabilitation of subacute stroke patients, especially for the recovery of hand motor function. RESULTS Clinical characteristics of the

sample at baseline are reported in Table 1. Two patients allocated to the Sonification Group interrupted the trial between T0 and T2: the first one worsened his/her clinical condition and

the second one was hospitalized. In addition, 18 patients in the Sonification Group and 16 patients in the Standard of Care Group were lost to follow-up due to the clinical worsening (2

patients in the Sonification Group and 3 patients in the Standard of Care Group) or to the difficulty to go back to the rehabilitative centers after the end of the interventions (16 patients

in the Sonification Group and 13 patients in the Standard of Care Group) (Flow Diagram is reported in Fig. 1). No side effects were observed in both the experimental and control groups

after training. PRIMARY OUTCOME The upper limb impairment assessed by the Fugl-Meyer Upper Extremity scale total score44 , showed a significantly greater improvement between T0 and T2 (Table

2, Fig. 2) in the Sonification Group compared to the Standard of Care Group (time*group interaction: p = 0.024), as pointed out by the Intention To Treat analysis. SECONDARY OUTCOMES At the

Intention To Treat analysis, other significantly greater improvements between T0 and T2 in the Sonification versus the Standard of Care Groups were found for the Distal Score (time*group

interaction: p = 0.016), Wrist Score (time*group interaction: p = 0.033) and Hand Score (time*group interaction: p = 0.04) subscales. Also, Manual dexterity significantly improved (T0 vs T2)

in the Sonification Group as showed by the Box and Block Test45 (affected limb time*group interaction: p = 0.020; unaffected limb time*group interaction: p = 0.004). The Box and Block Test

score Ratio improvement was not significant (time*group interaction: p = 0.19). In addition to motor outcomes the study put in evidence the effects of sonification in pain perception: the

Sonification Group showed a significant reduction in Numerical Pain Rating Scale46 scores compared to the Standard of Care Group (time*group interaction: p = 0.006). No significant changes

were observed in spasticity (Modified Ashworth Scale)47 and perceived quality of life (McGill Quality of Life)48. All results are summarized in Table 2 and Fig. 3 and the same findings were

confirmed at the Per Protocol analyses (Supplementary Table S1). DISCUSSION The study confirmed how the sonification with music elements is able to facilitate the recovery process compared

to the traditional rehabilitative approach. Main results are related to the global upper limb impairment (primary endpoint of the study), assessed by the Fugl-Meyer Upper Extremity scale

(FM-UE). In particular, both the experimental (SG) and control (SoCG) groups scored a clinically significant increase of the FM-UE total score after training, but only the SG had a

clinically significant (> 5) improvement49. Conversely to the control group the experimental group did not show a significant improvement at proximal level at T2. A possible explanation

could be that the Sonichand protocol involves mainly the distal segments of the arm and this aspect should be considered in future studies. Results of the present study can be extended to

manual dexterity, as testified by the improved Box and Block Test (BBT), both in affected and unaffected limbs. The score improvement observed for the presumed-to-be-unaffected upper

extremity could be due to a learning phenomenon obtained during repeated evaluations or to undetected motor deficit due to the brain lesion thus involving slight abnormal kinematics subject

to improvements50. In addition, some BBT variability could be due to a day-by-day effect that however should be minimized by taking into consideration the ratio score51. In accordance with

Kinney et al.52 thresholds, our study showed a large effect size for time effect (> 0.55) for both FM-UE total and subscales values and for the BBT of the affected upper limb, a medium

effect size (≥ 0.31, < 0.55) for BBT of unaffected upper limb and a small effect size (≥ 0.14, < 0.31) for the remaining outcome variables. Conversely to a study for robot assisted

hand training53, the effect size for group effect of our study was not statistically significant for all outcome variables. This could be likely due to the different training device and

intensity of training (20 min. vs. 60 min.) adopted in the two studies. Finally, small effect sizes of the time*group interactions should be evaluated in a framework of power of interactions

in the analysis of variance models. The sonification technique used in our study was very intuitive, compared to other similar interventions37. In fact, music support was intrinsically

associated with the patient's movements, regardless of his/her intention to produce sounds. This peculiarity allowed the patient to focus mainly on the movement and, at the same time,

to receive a natural feedback on its coherence and quality (i.e. a correct movement produced a fluent musical progression). The above mentioned mechanisms could have helped the patient to

model/adjust and re-learn movements during the rehabilitative intervention. Another important result was found by evaluating the effect of sonification on pain perception associated to upper

extremity movements during the rehabilitation intervention. Despite higher levels of perceived pain (Baseline scores) in the Sonification group compared to those perceived by the control

group patients, the music-based rehabilitative approach used in this study was able to produce a significant decrease of NPRS score. In accordance with the literature concerning the use of

music in pain medicine, our data showed how the presence of musical patterns can significantly reduce the perceived pain, likely because music helps patients to get distracted from the pain

condition54. The harmonic and melodic progressions used in the study, compared to other possible sounds (e.g. sinusoidal tones) could have contributed to create predictable and pleasant

sequences, involving the recall and recognition of the tonal trend typical of the music listening experience. In addition, the predictability of sequences could also have produced a

facilitation effect during the preparation and execution of movements. One of the main limitations of the study is the limited data available for the planned follow-up assessment at 1 month

after the end of treatment (T3). In spite of therapists’ recommendation, only less than 50% of patients completed the study, making the reliability of results at T3 questionable. This could

be due to the patients’ difficulties to get back to the rehabilitative centers after the end of the interventions and/or to other reasons, such as a lack of motivation/unwillingness due to a

follow-up too close to the study conclusion and conflicting with the new difficulties of daily life activities of the whole family. For this reason, we decided to include the follow-up

analysis only into the supplementary materials (Supplementary Table S2). This analysis showed a maintenance of results over time. Another limitation of the study is that we did not apply

block randomization in the allocation of patients to experimental and control groups. Taking into account the defined exclusion criteria, the block randomization would not have guaranteed in

a reasonable time, patients’ enrollment at the different Units, with a balanced and homogeneous sample size. In addition, an increased sample size could have produced greater effect size to

support the beneficial effect of the sonification technique. Finally, the lack of other evaluation methods, such as neuroimaging techniques (fMRI) aimed at pointing out possible changes in

the brain connectivity and their possible correlation with post-treatment performance indexes, should be considered as a further limitation. The study proved that this type of music

intervention can improve movements and reduce pain during the rehabilitation process in patients with stroke. Upper limb rehabilitation with sonification techniques can be considered a

standardized intervention, effective and easy to implement in subacute stroke rehabilitation. The low-cost of the sonification set-up makes the intervention widely exploitable into the

standard occupational/physical therapy settings. Also, it is expected that the Sonichand protocol and approach, thanks to the powerful feedback produced by movements sonification, could be

beneficial also for the treatment of other neurological conditions in which upper limbs movements are impaired due to specific brain lesions (i.e., gesture disturbances, apraxia and poor

gesture imitation). Future studies involving other clinical conditions and with a more comprehensive assessment protocol, could strengthen the significant findings obtained by this study.

METHODS This randomized controlled trial involved 65 subacute stroke patients who underwent inpatient rehabilitation allocated into two groups: the SG (n = 33) underwent sonification

sessions (see “Interventions” sub-section) and the SoCG (n = 32) underwent traditional motor rehabilitation sessions. The randomization was centrally managed: a randomization list (simple

randomization), generated accordance with a completely randomized parallel group study design, was generated at the beginning of the study and delivered to the Principal Investigator (PI) of

the study. Each recruited subject was allocated in SG or SoCG by the PI on the basis of the randomization list by assigning a unique identifying code so to guarantee data anonymization

during data collection, analysis and interpretation processes. The blindness of the evaluations was guaranteed both for questionnaires data collection and statistical processing. Subjects

were recruited from March 27th 2017 to April 15th 2019 in 5 Italian rehabilitative units located at Maugeri Scientific Clinical Institutes of Pavia, Montescano (PV) and Nervi (GE), S. Lucia

Foundation IRCCS (Rome) and Neurological Clinic of S. Martino Hospital (University of Genoa). All procedures performed in this study were in accordance with the ethical standards of the

institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. The study was approved by the Ethical Board

of the Istituti Clinici Scientifici Maugeri IRCCS, Pavia, Italy (2088 CE, December 19th, 2016) and registered on https://clinicaltrials.gov (Trial Number Identifier: NCT03306797, date of

registration: 11/10/2017), all participants gave their written informed consent to participate in the study. INCLUSION AND EXCLUSION CRITERIA Patients matching the following characteristics

were included in the study: * Age 40–85 years * Ischaemic lesion in a single hemisphere (right or left hemiplegia/hemiparesis) * Clinically evaluable residual movement capacity of the

paretic upper limb (ability to autonomously make postural adjustments during reaching tasks) * Mini Mental State Examination > 24 * Acute onset no more than 180 days prior to enrollment

in the study The exclusion criteria of the Study were the following: * Presence of neglect (evaluated through a clinical and functional assessment) * previous or concomitant diseases

affecting upper limb functions (e.g. Parkinson's disease, multiple sclerosis, shoulder periarthritis, Dupuytren's disease, etc.) * rehabilitation treatments with music in the last

year INTERVENTIONS The patients recruited in the study were admitted to the rehabilitation centers and received usual treatment for their subacute stage of disease (such as occupational

therapy, speech therapy, psychological support, lower extremity rehabilitation, etc.) in addition to a specific upper extremity rehabilitation based on (a) standard motor exercises or (b)

standard motor exercises supported by sonification techniques. STANDARD MOTOR EXERCISES The upper limb rehabilitation was carried out in accordance with the Italian Stroke Prevention and

Educational Awareness Diffusion (SPREAD) guidelines55 and took into account a traditional neuro-motor approach aimed at improving motor control and reducing spasticity. At the same time,

these methods were combined with an individualized and progressive functional re-education through task-oriented programs56. The first phase of the treatment (passive treatment, 15 min)

aimed at relaxing muscle tone with stretching. The exercises were chosen, based on the needs and residual motor skills of the patient, from among the following: * fingers opening and closing

* prono-supination of the forearm * wrist mobilization The second phase of the treatment (20 min) included movements actively performed by the patient. The patient was offered at least 6

different exercises. Each exercise lasted 90 s and was followed by a break of 30–60 s. The exercises were chosen from the following: * 1. Forearm–wrist: (a) flexion and extension of the

wrist in the intermediate position of the forearm (horizontal plane), (b) flexion and extension of the wrist with pronated arm (vertical plane), (c) ulnar and radial deviation of the wrist,

(d) pronation and supination of the forearm. * 2. Hand: (a) grasp; (b) pinch; (c) finger extension (d) abduction and adduction of the fingers * 3. Shoulder-elbow: (a) elbow flexion and

extension; (b) combined movement of shoulder flexion and elbow extension During the active phase the therapist assisted the patient in the movement that took place autonomously, with the

active participation of the patient. The exploration of the workspace was carried out using tools such as a table facilitating sliding of the arm on the work surface. If compatible with

patient’s impairment and performance, grip-release exercises of medium-volume objects and their displacement in the space were also included. STANDARD MOTOR EXERCISES SUPPORTED BY

SONIFICATION TECHNIQUES This intervention was similar to the above mentioned treatment (a), i.e. it included a passive treatment phase without sonification (15 min) and a second phase (20 

min) in which the motor exercises were supported by sonification techniques. In this second phase of treatment the patient performed the movements actively, by practicing a sequence of

exercises lasting about 90 s each, separated by a 30–60 s resting period. The sonification process was made using an adequate setup including a technological sensor (Leap Motion

Controller)57,58,59, managed with an ad-hoc developed application parameterizing the data read by the sensor by producing sounds in response to some specific hand movements. The Leap Motion

Controller (LMC; Leap Motion, Inc., San Francisco, CA, USA) is described as a low-cost Human Computer Interface specifically optimized for the recognition of gestures produced by hands,

fingers and pointer objects, with a precision, validated in real situation, of about 0.7 mm. The LMC is connected to the computer through a USB 3.0 communication interface and does not

require contact with the patient. Thanks to the use of 3 infrared LEDs and 2 optical sensors, it can create a three-dimensional workspace of about 120 cm × 60 cm × 60 cm. The LMC is capable

to capture the movement of hands and fingers thanks to two monochromatic cameras and three infrared LEDs and it is able to build a vector representation of the hand, fingers, wrist and

forearm up to the elbow. The device is not so precise for temporal responses but, unlike tapping, this type of exercises do not need a strict temporal association between movements and

sounds. Before starting, the minimum and maximum points of the movements (range of motion) are measured. If the sonification is discrete (e.g. 4 notes of a musical scale) this range is

further divided into thresholds. During the performance, if the read value falls within a preset range around a minimum/maximum threshold, a preset sound sample is played (basically each

threshold is a trigger for the assigned note). If the sonification is continuous, the read value will modulate the sound along a continuum, from the minimum to the maximum level measured

(range of motion). In this trial we used a sampling rate of 25 Hz. The sensor, which is positioned under the palm of the hand at a distance of about 15 cm, identifies and maps the limb,

creating a representation of it in points, providing the relative kinetic characteristics (spatial coordinates, acceleration, direction, inclination, etc.) for each point. The software

program (https://ismm.ircam.fr/leapmotion/ ), used to carry out our experiments, was developed with the Max 7 software programming environment (Max 7 by Cycling ’74, Walnut, CA, USA,

https://cycling74.com). Thanks to this application the patient's gestures are supported by two modalities: * 1. In the first, movements (forearm-wrist: flexion and extension of the

wrist in the intermediate position of the forearm—horizontal plane-, flexion and extension of the wrist with pronated arm—vertical plane-, ulnar and radial deviation of the wrist; hand:

pinch, finger extension; shoulder-elbow: elbow flexion and extension) trigger and modulate a harmonic progression built on the consecutive grades of the major scale played as an ascending

and descending arpeggio also with a consistent volume (ascending crescendo and descending diminuendo sounds) (Supplementary Video S3). * 2. In the second, movements (forearm-wrist: pronation

and supination of the forearm; hand: grasp, abduction and adduction of the fingers; shoulder-elbow: combined movement of shoulder flexion and elbow extension) modulate the volume and

low-pass cutoff frequency of a tone filter applied to a synthetic texture. It is a kind of tone modulation of a pad sound, from dull to bright, based on a harmonic progression built on the

consecutive grades of a major scale (Supplementary Video S4). On the basis of the exercises selected for rehabilitation, some music parameters (pitch, volume and spectrum—quality/brilliance

of tone) have been identified and superimposed on the gesture. The correct execution of the gesture leads patients to cyclically perform a harmonic progression of arpeggios. Considering the

daily administration of the training protocol, different instruments were associated to the various exercises to increase the sound variability (for example: trumpet, flute, cello, guitar,

piano etc., as well as some timbre variants for the pad performed by a synthesizer). To ensure correct understanding of the training protocol, before the experimental intervention, the

therapist measured the active range of motion for sensor mapping and provided a short demonstration of the exercises (learning phase) to the patient who carried out a short training with the

non-plegic limb. In addition, in order to guarantee uniformity of intervention, the experimental set up (including a wheelchair with a transparent table and a clamp to support the Leap

Motion Controller) was replicated in each Unit involved in the study. All interventions (5 days a week for 4 weeks, for a total of 20 sessions) were assisted by physiotherapists or

occupational therapists with an expertise in hand rehabilitation or neurorehabilitation and specifically trained to use the Leap Motion Controller and the sonification described setting. The

therapists involved into the protocol received at least 8 h of specific training before the beginning of experiments. ASSESSMENT The rehabilitation interventions were evaluated by blinded

evaluators at baseline (T0), after 2 weeks (T1), at the end of treatment (T2, after 4 weeks) and 1 month after the end of the treatment (T3). The FM-UE44 (score range 0–66) was used to

evaluate the performance-based sensorimotor functions of the impaired upper limb. Furthermore, in order to detail the motor improvement both at the proximal and distal level, the FM-UE was

sub-scaled into proximal (shoulder/elbow, score range 0–42), distal (wrist/hand, score range 0–24), wrist (score range 0–10) and hand (score range 0–14) subscores60. The Modified Ashworth

Scale (MAS, score range 0–4; 6 categories) was used to evaluate muscle tone at the upper limb61. In accordance with previous studies, the score of ‘1+’ category was converted to 1.4 to be

managed in statistical analysis60. The BBT was used as a measure of gross manual dexterity62. In particular, for this test we assessed the performance (number of blocks) both of the affected

and unaffected limbs and their score ratio (BBT ratio = BBT affected limb/BBT unaffected limb) in order to compensate for their variability51. Pain and quality of life assessment were rated

using respectively the NPRS63 (score range 0–10) and the McGill Quality of Life (MQOL-it, score range 0–160)48. OUTCOMES MEASURES The primary outcome measure of the study was the evaluation

of the improvement in total scores of the FM-UE between T0 and T2. Secondary outcomes measure were other motor aspects (including scores of BBT, MAS and sub-items scores of the FM-UE), pain

(scores of NPRS) and the perceived quality of life (MQOL-it). STATISTICS Sample size: the sample size estimate is based on previous data acquired from results of rehabilitation

interventions, with and without technology-assisted techniques, carried out at the facilities participating in the study64. For the sizing of the present study it is assumed that the use of

a technological system for sonification allows to achieve at least clinically relevant results as traditional rehabilitation. An increase of at least 6 points in FM-UE is recognized to be

the minimal clinically important difference between baseline and end of treatment following rehabilitation54. The work hypothesis, based on previous data, was that the pre-post treatment

difference in FM-UE was at least 5 point higher in the SG than in the SoCG groups. With these hypotheses and assuming a standard deviation of the pre-post treatment differences of 10, a

total sample of 65 subjects can guarantee a study power of 80% and a type I error of 5%. DATA ANALYSIS Descriptive statistics were carried out for all the collected variables. Means and

standard deviations were reported for normally distributed quantitative variables. In case of significant discrepancy from the normality, medians and interquartile ranges were also included.

Nominal level variables were summarized as frequencies and percentages. Baseline clinical characteristics between SG and SoG groups were compared by means of Student’s t test, Mann–Whitney

U test or chi-square test according to the nature and distribution of the variables. Inferential statistics were performed on the Intention To Treat population defined as all randomized

subjects. Handling of missing data was managed according to a strategy pre-specified in the study protocol and referred to the four time points of assessment. Imputation strategy included

the handling of missing at random and not at random. Generalized linear mixed models were applied to handle missingness at single timepoints where measured repeatedly over time. Last

Observation Carried Forward (LOCF) method was applied to missingness of patients who prematurely interrupted the trial. Analyses on the PP population were also conducted to support the

findings. Repeated measures analysis of variance with one factor (factor: experimental/control group) or non-parametric approach (Kruskall-Wallis statistics) was used to assess potential

differences between baseline (T0) and end of treatment (T2) for the primary and secondary endpoints. The same models were applied to assess trends over time at the four assessment timepoints

foreseen in the protocol (T0 = baseline T1 = 2 weeks of treatment, T2 = 4 weeks, end of treatment, T3 = 8 weeks, follow-up). Due to the huge amount of missing data at the follow-up (T3),

this analysis was carried out on the per protocol population only without any imputation of missing data. Effect sizes of the main effects, repeated measures and interactions were finally

computed to point out the magnitude of the effects in a standardized metric with the aim of favoring comparisons with other studies. Effect size was rated as small (d = 0.14), medium (d = 

0.31) and large (d = 0.55) in accordance with the effect size interpretation guidelines provided by Kinney and coll. for a study classified as a Multicomponent intervention52.

Experiment-wise error rate was set at 5%. All analyses were conducted using SPSS Statistics Professional package. DATA AVAILABILITY The datasets generated during and/or analyzed during the

current study are available from the corresponding author on reasonable request. REFERENCES * Langhorne, P., Coupar, F. & Pollock, A. Motor recovery after stroke: A systematic review.

_Lancet Neurol._ 8, 741–754 (2009). Article  PubMed  Google Scholar  * Pollock, A. _et al._ Interventions for improving upper limb function after stroke. _Cochrane Database Syst. Rev._ 12,

CD0102820 (2014). Google Scholar  * Kwakkel, G. & Kollen, B. J. Predicting activities after stroke: What is clinically relevant?. _Int. J. Stroke_ 8, 25–32 (2013). Article  CAS  PubMed 

Google Scholar  * Nowak, D. A., Hermsdörfer, J. & Topka, H. Deficits of predictive grip force control during object manipulation in acute stroke. _J. Neurol._ 250, 850–860 (2003).

Article  PubMed  Google Scholar  * Kwakkel, G. Impact of intensity of practice after stroke: Issues for consideration. _Disabil. Rehabil._ 28, 823–830 (2006). Article  PubMed  Google Scholar

  * Maclean, N., Pound, P., Wolfe, C. & Rudd, A. Qualitative analysis of stroke patients’ motivation for rehabilitation. _BMJ_ 321, 1051–1054 (2000). Article  CAS  PubMed  PubMed Central

  Google Scholar  * Iosa, M. _et al._ Seven capital devices for the future of stroke rehabilitation. _Stroke Res. Treat._ 2012, 187965 (2012). CAS  PubMed  PubMed Central  Google Scholar  *

Moens, B. & Leman, M. Alignment strategies for the entrainment of music and movement rhythms. _Ann. N. Y. Acad. Sci._ 1337, 86–93 (2015). Article  ADS  PubMed  Google Scholar  * Morone,

G. _et al._ Rehabilitative devices for a top-down approach. _Expert Rev. Med. Devices_ 16, 187–195 (2019). Article  CAS  PubMed  Google Scholar  * Dauvergne, C. _et al._ Home-based training

of rhythmic skills with a serious game in Parkinson’s disease: Usability and acceptability. _Ann. Phys. Rehabil. Med._ 61, 380–385 (2018). Article  PubMed  Google Scholar  * Mehrholz, J.,

Pohl, M., Platz, T., Kugler, J. & Elsner, B. Electromechanical and robot-assisted arm training for improving activities of daily living, arm function, and arm muscle strength after

stroke. _Cochrane Database Syst. Rev._ 9, CD006876 (2018). PubMed  Google Scholar  * Balasubramanian, S., Klein, J. & Burdet, E. Robot-assisted rehabilitation of hand function. _Curr.

Opin. Neurol._ 23, 661–670 (2010). Article  PubMed  Google Scholar  * Yue, Z., Zhang, X. & Wang, J. Hand rehabilitation robotics on poststroke motor recovery. _Behav. Neurol._ 2017,

3908135 (2017). Article  PubMed  PubMed Central  Google Scholar  * Moumdjian, L., Sarkamo, T., Leone, C., Leman, M. & Feys, P. Effectiveness of music-based interventions on motricity or

cognitive functioning in neurological populations: A systematic review. _Eur. J. Phys. Rehabil. Med._ 53, 466–482 (2017). PubMed  Google Scholar  * Schlaug, G. Part VI introduction:

Listening to and making music facilitates brain recovery processes. _Ann. N. Y. Acad. Sci._ 1169, 372–373 (2009). Article  ADS  PubMed  PubMed Central  Google Scholar  * Altenmuller, E.,

Marco-Pallares, J., Munte, T. F. & Schneider, S. Neural reorganization underlies improvement in stroke-induced motor dysfunction by music-supported therapy. _Ann. N. Y. Acad. Sci._ 1169,

395–405 (2009). Article  ADS  CAS  PubMed  Google Scholar  * Ripollés, P. _et al._ Music supported therapy promotes motor plasticity in individuals with chronic stroke. _Brain Imaging

Behav._ 10, 1289–1307 (2016). Article  PubMed  Google Scholar  * Sihvonen, A. J. _et al._ Music-based interventions in neurological rehabilitation. _Lancet Neurol._ 16, 648–660 (2017).

Article  PubMed  Google Scholar  * Herholz, S. C. & Zatorre, R. J. Musical training as a framework for brain plasticity: Behavior, function, and structure. _Neuron_ 76, 486–502 (2012).

Article  CAS  PubMed  Google Scholar  * Bangert, M. & Altenmuller, E. Mapping perception to action in piano practice: A longitudinal DC-EEG-study. _BMC Neurosci._ 4, 26–36 (2003).

Article  PubMed  PubMed Central  Google Scholar  * Bangert, M. _et al._ Shared networks for auditory and motor processing in professional pianists: Evidence from fMRI conjunction.

_Neuroimage_ 30, 917–926 (2006). Article  PubMed  Google Scholar  * Suh, J. H. _et al._ Effect of rhythmic auditory stimulation on gait and balance in hemiplegic stroke patients.

_NeuroRehabilitation_ 34, 193–199 (2014). Article  PubMed  Google Scholar  * Hayden, R., Clair, A. A., Johnson, G. & Otto, D. The effect of rhythmic auditory stimulation (RAS) on

physical therapy outcomes for patients in gait training following stroke: A feasibility study. _Int. J. Neurosci._ 119, 2183–2195 (2009). Article  PubMed  Google Scholar  * van Wijck, F. _et

al._ Making music after stroke: Using musical activities to enhance arm function. _Ann. N. Y. Acad. Sci._ 1252, 305–311 (2012). Article  ADS  PubMed  Google Scholar  * Schneider, S.,

Schonle, P. W., Altenmuller, E. & Munte, T. F. Using musical instruments to improve motor skill recovery following a stroke. _J. Neurol._ 254, 1339–1346 (2007). Article  CAS  PubMed 

Google Scholar  * Lim, K. B. _et al._ The therapeutic effect of neurologic music therapy and speech language therapy in post-stroke aphasic patients. _Ann. Rehabil. Med._ 37, 556–562 (2013).

Article  ADS  PubMed  PubMed Central  Google Scholar  * Kim, S. J. & Jo, U. Study of accent-based music speech protocol development for improving voice problems in stroke patients with

mixed dysarthria. _NeuroRehabilitation_ 32, 185–190 (2013). Article  PubMed  Google Scholar  * Jun, E. M., Roh, Y. H. & Kim, M. J. The effect of music-movement therapy on physical and

psychological states of stroke patients. _J. Clin. Nurs._ 22, 22–31 (2013). Article  PubMed  Google Scholar  * Kim, D. S. _et al._ Effects of music therapy on mood in stroke patients.

_Yonsei Med. J._ 52, 977–981 (2011). Article  ADS  PubMed  PubMed Central  Google Scholar  * Magee, W. L. & Davidson, J. W. The effect of music therapy on mood states in neurological

patients: A pilot study. _J. Music Ther._ 39, 20–29 (2002). Article  PubMed  Google Scholar  * Bevilacqua, F. _et al._ Sensori-motor learning with movement sonification: Perspectives from

recent interdisciplinary studies. _Front. Neurosci._ 10, 385 (2016). Article  PubMed  PubMed Central  Google Scholar  * Effenberg, O., Fehse, U., Schmitz, G., Krueger, B. & Mechling, H.

Movement sonification: Effects on motor learning beyond rhythmic adjustments. _Front. Neurosci._ 10, 219 (2016). Article  PubMed  PubMed Central  Google Scholar  * Friedman, N. _et al._

Retraining and assessing hand movement after stroke using the MusicGlove: Comparison with conventional hand therapy and isometric grip training. _J. Neuroeng. Rehabil._ 11, 76 (2014).

Article  PubMed  PubMed Central  Google Scholar  * Zondervan, D. K. _et al._ Home-based hand rehabilitation after chronic stroke: Randomized, controlled single-blind trial comparing the

MusicGlove with a conventional exercise program. _J. Rehabil. Res. Dev._ 53, 457–472 (2016). Article  PubMed  Google Scholar  * Scholz, D. S. _et al._ Sonification as a possible stroke

rehabilitation strategy. _Front. Neurosci._ 8, 332 (2014). Article  PubMed  PubMed Central  Google Scholar  * Scholz, D. S., Rhode, S., Großbach, M., Rollnik, J. & Altenmüller, E. Moving

with music for stroke rehabilitation: A sonification feasibility study. _Ann. N. Y. Acad. Sci._ 1337, 69–76 (2015). Article  ADS  PubMed  Google Scholar  * Scholz, D. S. _et al._

Sonification of arm movements in stroke rehabilitation—a novel approach in neurologic music therapy. _Front. Neurol._ 7, 106 (2016). Article  PubMed  PubMed Central  Google Scholar  *

Schmitz, G. _et al._ Movement sonification in stroke rehabilitation. _Front. Neurol._ 9, 389 (2018). Article  PubMed  PubMed Central  Google Scholar  * Ghai, S., Schmitz, G., Hwang, T. H.

& Effenberg, A. O. Auditory proprioceptive integration: Effects of real-time kinematic auditory feedback on knee proprioception. _Front. Neurosci._ 12, 142 (2018). Article  PubMed 

PubMed Central  Google Scholar  * Effenberg, A. O. & Schmitz, G. Acceleration and deceleration at constant speed: Systematic modulation of motion perception by kinematic sonification.

_Ann. N. Y. Acad. Sci._ 1425, 52–69 (2018). Article  ADS  PubMed  Google Scholar  * Nikmaram, N. _et al._ Musical sonification of arm movements in stroke rehabilitation yields limited

benefits. _Front. Neurosci._ 13, 1378 (2019). Article  PubMed  PubMed Central  Google Scholar  * Reh, J., Hwang, T. H., Schmitz, G. & Effenberg, A. O. Dual mode gait sonification for

rehabilitation after unilateral hip arthroplasty. _Brain Sci._ 9, 66 (2019). Article  PubMed Central  Google Scholar  * Colombo, R. _et al._ The SonicHand protocol for rehabilitation of hand

motor function: A validation and feasibility study. _IEEE Trans. Neural. Syst. Rehabil. Eng._ 27, 664–672 (2019). Article  CAS  PubMed  Google Scholar  * Chen, P., Lai, C. K., Chung, R. C.

& Ng, S. S. The Jacket Test for assessing people with chronic stroke. _Disabil. Rehabil._ 39, 2577–2583 (2017). Article  PubMed  Google Scholar  * Leathley, M. J. _et al._ Reliability of

measurements of muscle tone and muscle power in stroke patients. _Age Ageing_ 29, 223–228 (2000). Article  PubMed  Google Scholar  * Desrosiers, J., Bravo, G., Hébert, R., Dutil, E. &

Mercier, L. Validation of the box and block test as a measure of dexerity of elderly people: Reliability, validity, and norms studies. _Arch. Phys. Med. Rehabil._ 75, 751–755 (1994). Article

  CAS  PubMed  Google Scholar  * McCaffery, M. & Beebe, A. _Pain: Clinical Manual for Nursing Practice_ (CV Mosby Company, 1989). Google Scholar  * Sguazzin, C., Giorgi, I., Alesii, A.

& Fini, M. Italian validation of the McGill Quality of Life Questionnaire (MQOL-It). _G. Ital. Med. Lav. Ergon._ 32, B58-62 (2010). PubMed  Google Scholar  * Page, S. J., Fulk, G. D.

& Boyne, P. Clinically important differences for the upper-extremity Fugl-Meyer Scale in people with minimal to moderate impairment due to chronic stroke. _Phys. Ther._ 92, 791–798

(2012). Article  PubMed  Google Scholar  * Nakamura, T., Abreu, B. C., Patterson, R. M., Buford, W. L. Jr. & Ottenbacher, K. J. Upper-limb kinematics of the presumed-to-be-unaffected

side after brain injury. _Am. J. Occup. Ther._ 62, 46–50 (2008). Article  PubMed  Google Scholar  * Mathiowetz, V., Volland, G., Kashman, N. & Weber, K. Adult norms for the Box and Block

Test of manual dexterity. _Am. J. Occup. Ther._ 39, 386–391 (1985). Article  CAS  PubMed  Google Scholar  * Kinney, A. R., Eakman, A. M. & Graham, J. E. Novel effect size interpretation

guidelines and an evaluation of statistical power in rehabilitation research. _Arch. Phys. Med. Rehabil._ 101, 2219–2226 (2020). Article  PubMed  Google Scholar  * Orihuela-Espina, F. _et

al._ Robot training for hand motor recovery in subacute stroke patients: A randomized controlled trial. _J. Hand Ther._ 29, 51–57 (2016). Article  PubMed  Google Scholar  * Linnemann, A. _et

al._ The effects of music listening on pain and stress in the daily life of patients with fibromyalgia syndrome. _Front. Hum. Neurosci._ 9, 434 (2015). Article  PubMed  PubMed Central 

Google Scholar  * AA.VV. Trattamento riabilitativo e continuità dell’assistenza, in Stroke Prevention And Educational Awareness Diffusion (SPREAD), Ictus Cerebrale: Linee guida italiane di

prevenzione e trattamento, Pubblicazioni Catel-Hyperphar Group SpA, pp. 297–376 (2003). * Carr, J. H. & Shepherd, R. B. A motor learning model for stroke rehabilitation. _Physiotherapy_

75, 372–380 (1989). Article  Google Scholar  * Iosa, M. _et al._ Leap motion controller videogame-based therapy for rehabilitation of elderly patients with post-acute stroke: A feasibility

pilot study. _Top Stroke Rehabil._ 22, 306–316 (2015). Article  PubMed  Google Scholar  * Smeragliuolo, A. H., Hill, N. J., Disla, L. & Putrino, D. Validation of the leap motion

controller using markered motion capture technology. _J. Biomech._ 49, 1742–1750 (2016). Article  PubMed  Google Scholar  * Li, W., Hsieh, C., Lin, L. & Chu, W. Hand gesture recognition

for post-stroke rehabilitation using leap motion. _Int. Conf. Appl. Syst. Innov._ 2017, 386–388 (2017). Google Scholar  * Qian, Q. _et al._ Distal versus proximal—an investigation on

different supportive strategies by robots for upper limb rehabilitation after stroke: A randomized controlled trial. _J. Neuroeng. Rehabil._ 3, 64 (2019). Article  Google Scholar  *

Bohannon, R. W. & Smith, M. B. Interrater reliability of a modified Ashworth scale of muscle spasticity. _Phys. Ther._ 67, 206–207 (1987). Article  CAS  PubMed  Google Scholar  * Platz,

T. _et al._ Reliability and validity of arm function assessment with standardized guidelines for the Fugl-Meyer Test, Action Research Arm Test and Box and Block Test: A multicentre study.

_Clin. Rehabil._ 19, 404–411 (2005). Article  PubMed  Google Scholar  * Williamson, A. & Hoggart, B. Pain: A review of three commonly used pain rating scales. _J. Clin. Nurs._ 14,

798–804 (2005). Article  PubMed  Google Scholar  * Colombo, R. _et al._ Comparison of exercise training effect with different robotic devices for upper limb rehabilitation: A retrospective

study. _Eur. J. Phys. Rehabil. Med._ 53, 240–248 (2017). PubMed  Google Scholar  Download references FUNDING This work was partially supported by the “Ricerca Corrente” funding provided by

the Italian Ministry of Health. AUTHOR INFORMATION AUTHORS AND AFFILIATIONS * Istituti Clinici Scientifici Maugeri, I.R.C.C.S., Istituti Clinici Scientifici Maugeri, Music Therapy Research

Laboratory, Scientific Institute of Pavia , Via Maugeri 10, 27100, Pavia, Italy Alfredo Raglio, Monica Panigazzi, Roberto Colombo, Paola Baiardi, Daniele Molteni, Chiara Imbriani, Chiara

Imarisio & Marcello Imbriani * Fondazione S. Lucia, I.R.C.C.S., Rome, Italy Marco Tramontano, Marco Iosa, Sara Mastrogiacomo & Carlo Caltagirone * Istituti Clinici Scientifici

Maugeri, I.R.C.C.S., Nervi (GE), Pavia, Italy Eleonora Baldissarro & Caterina Pistarini * Neurological Clinic, S. Martino Hospital, University of Genoa, Genoa, Italy Mehrnaz Hamedani 

& Gian Luigi Mancardi * Istituti Clinici Scientifici Maugeri, I.R.C.C.S, Montescano, PV, Italy Laura Eretti Authors * Alfredo Raglio View author publications You can also search for this

author inPubMed Google Scholar * Monica Panigazzi View author publications You can also search for this author inPubMed Google Scholar * Roberto Colombo View author publications You can

also search for this author inPubMed Google Scholar * Marco Tramontano View author publications You can also search for this author inPubMed Google Scholar * Marco Iosa View author

publications You can also search for this author inPubMed Google Scholar * Sara Mastrogiacomo View author publications You can also search for this author inPubMed Google Scholar * Paola

Baiardi View author publications You can also search for this author inPubMed Google Scholar * Daniele Molteni View author publications You can also search for this author inPubMed Google

Scholar * Eleonora Baldissarro View author publications You can also search for this author inPubMed Google Scholar * Chiara Imbriani View author publications You can also search for this

author inPubMed Google Scholar * Chiara Imarisio View author publications You can also search for this author inPubMed Google Scholar * Laura Eretti View author publications You can also

search for this author inPubMed Google Scholar * Mehrnaz Hamedani View author publications You can also search for this author inPubMed Google Scholar * Caterina Pistarini View author

publications You can also search for this author inPubMed Google Scholar * Marcello Imbriani View author publications You can also search for this author inPubMed Google Scholar * Gian Luigi

Mancardi View author publications You can also search for this author inPubMed Google Scholar * Carlo Caltagirone View author publications You can also search for this author inPubMed 

Google Scholar CONTRIBUTIONS A.R. designed the experiment and the protocol, provided music therapy expertise, interpreted the results, and prepared the manuscript. S.M., C.I., L.E. and M.H.

performed the experiments. M.P., M.T., E.B. and C.P. provided clinical support and supervision. P.B. performed statistical analysis. R.C. and M.I. provided engineering support and

contributed to prepare the manuscript. D.M. and C.I. provided music and music technology expertise. M.I., G.M. and C.C. contributed to data interpretation and critical reading/revision of

the manuscript. All authors have read and approve of the final manuscript version. CORRESPONDING AUTHOR Correspondence to Alfredo Raglio. ETHICS DECLARATIONS COMPETING INTERESTS The authors

declare no competing interests. ADDITIONAL INFORMATION PUBLISHER'S NOTE Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional

affiliations. SUPPLEMENTARY INFORMATION SUPPLEMENTARY INFORMATION 1. SUPPLEMENTARY INFORMATION 2. Supplementary Video 1. Supplementary Video 2. RIGHTS AND PERMISSIONS OPEN ACCESS This

article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as

you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party

material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the

article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the

copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. Reprints and permissions ABOUT THIS ARTICLE CITE THIS ARTICLE Raglio, A., Panigazzi, M.,

Colombo, R. _et al._ Hand rehabilitation with sonification techniques in the subacute stage of stroke. _Sci Rep_ 11, 7237 (2021). https://doi.org/10.1038/s41598-021-86627-y Download citation

* Received: 03 September 2020 * Accepted: 17 March 2021 * Published: 31 March 2021 * DOI: https://doi.org/10.1038/s41598-021-86627-y SHARE THIS ARTICLE Anyone you share the following link

with will be able to read this content: Get shareable link Sorry, a shareable link is not currently available for this article. Copy to clipboard Provided by the Springer Nature SharedIt

content-sharing initiative