RETRACTED: Assessment Capacity of the Armeo^® Power: Cross-Sectional Study

Abstract

The use of robotics in rehabilitating motor functions has increased exponentially in recent decades. One of the most used robotic tools is undoubtedly the Armeo^® Power, which has proved to have excellent qualities as a rehabilitation tool. However, none of these studies has investigated the ability of Armeo^® Power to assess the upper limb by correlating the data resulting from the software with patient-reported outcome measures (PROMs). The present study aims to evaluate the variability between the standardized PROMs, Stroke Upper Limb Capacity Scale (SULCS), Fugl–Meyer upper limb assessment (FMA-UL), and the Armeo^® Power measurements. To evaluate the correlation between SULCS and FMA-UL and the strength and joint assessments obtained with the Armeo^® Power, Pearson’s correlation coefficient was used. A total of 102 stroke survivors were included in this cross-sectional study, and all participants finished the study. The results showed many statistically significant correlations between PROM items and Armeo^® Power data. In conclusion, from this study, it can be stated that Armeo^® Power, based on the analysis of the data collected, can be an objective evaluation tool, which can be combined with the operator-employee traditional evaluation techniques, especially when compared to a patient-reported outcome measures (PROMs).

1. Introduction

Worldwide, 12.2 million people suffer stroke annually, and of these 5 million are in great need of rehabilitation [1]. These numbers are expected to increase over the next century due to a growing aging population [2].

Most stroke patients survive the initial injury; however, long-term disabilities affecting the activities of daily living and participation are common [3]. Along with lesion location and size, rehabilitation activities can affect the degree of recovery [3]. Therefore, rehabilitation after a stroke is of utmost importance to alleviate and regain the lost functions. Importantly, upper limb paresis has been reported to appear in around 70% of stroke survivors [4], and regaining this function is essential in stroke rehabilitation to gain independence in activities of daily living (ADL). Rehabilitation services are the primary mechanism by which functional recovery and the achievement of independence are promoted in patients with acute stroke. Ideally, rehabilitation services are delivered by a multidisciplinary team of healthcare providers with training in neurology, rehabilitation nursing, occupational therapy, physical therapy, and speech and language therapy (SLT). Such teams are directed by physicians trained in physical medicine and rehabilitation (physiatrists) or by neurologists who have specialized training or board certification in rehabilitation medicine. Other health professionals who play an essential role in the process include social workers, psychologists, psychiatrists, and counselors [5,6].

Moreover, the use of robotics in rehabilitating motor functions has increased exponentially in recent decades [7]. Advances in non-invasive robotic-assisted rehabilitation promise to augment and facilitate these mechanisms during recovery after stroke. Several studies have shown that robot-assisted exercise can ensure more repetitions within a therapy session, compared to conventional therapy, suggesting superior effectiveness [8]. However, the discrepancy in how these protocols are applied and reported makes interpretation and generalization regarding results difficult. Numerous systematic reviews have investigated the effectiveness of robotics over conventional therapies, all confirming the need for further recommendations about the duration of a single session (number of minutes), the duration of the intervention period (number of weeks), the frequency (times a week), and intensity (energy expenditure) [9,10,11,12].

Moreover, numerous studies have investigated the effectiveness of robot-assisted hand and arm training compared with conventional therapy [13,14]. The results indicate beneficial effects on upper limb recovery, strength, motor control, and ADL when rehabilitation makes use of a robot [15,16,17]. However, an important aspect missing in clinical studies is the ability to assess these tools; in fact, most of them provide the clinician with a precise evaluation of the range of movement (ROM), strength, speed, and fluidity of movement, but no studies have investigated the correlation between these data and patient-reported outcome measures (PROMs). These tools in fact are used both in clinical practice and research settings to assess patient characteristics before any intervention and determine if patients have made meaningful changes in their recovery process, and may influence the intensity and duration of care [18]. Researchers need assessment tools such as these during the investigation of the efficacy and effectiveness of a given treatment intervention in order to have an objective and quantitative evaluation, supported by a subjective evaluation by the clinician and self-evaluation by the patient. One of the most used robotic tools is undoubtedly the Armeo^® Power, a robotic rehabilitation exoskeleton for upper limb training. Providing intelligent arm support in a 3D workspace, the Armeo^® Power has proved to have excellent qualities as a rehabilitation tool [19,20,21,22,23,24]. However, none of these studies have investigated the ability of Armeo^® Power to assess the upper limb by correlating the data resulting from the software with PROMs. Further investigation of Armeo^® Power’s assessment capacity would benefit patients, researchers, and clinicians. In fact, it is important for the development of clinical practice and research that practical and appropriate measures are universally accepted; this would allow comparisons and meta-analyses of high-quality randomized controlled trials. Therefore, the present study aimed to evaluate the variability between the standardized PROMs Stroke Upper Limb Capacity Scale (SULCS) [25], Fugl–Meyer upper limb assessment (FMA-UL) [26], and the Armeo^® Power measurements. This would allow for comparing objective assessments with subjective evaluations administered by the clinicians and self-evaluations provided by the patient.

2. Methods

2.1. Participants

For this cross-sectional study, participants were recruited at the neuromotor department of IRCCS San Raffaele Pisana, hospitalized in the period between July 2022 and September 2022. Study participants had to meet the following inclusion criteria: stroke diagnosis confirmed by imaging tests [27], age > 18 years old, period of hospitalization after the acute event of 3 weeks, absence of severely disabling comorbidity, and Ashworth scales < 4. All eligible individuals were informed about the methods and procedures of the study, and all interested persons signed an informed consent form. Information about each patient, such as diagnosis and treatment, was culled from medical records and filled in by the treating physician.

2.2. Tools

The Armeo^® Power is a 6-degrees-of-freedom (https://www.hocoma.com/solutions/armeo-power/, accessed on 2 August 2023) exoskeleton that allows the rehabilitation of the hemiplegic limb in a stroke patient. The device can support the weight of the patient’s arm, thus providing a floating sensation, and assists him in a large 3D workspace while performing the exercises. The presence of a suspension system allows the facilitator to set and adjust the robot’s sensitivity according to each patient’s characteristics. The arm and forearm lengths are both adjustable so that the device can be adapted for use by a wide selection of patients [21].

This machine is intended for patients who have lost or impaired function in the upper limbs due to neurogenic, spinal, muscle, or bone-related central or peripheral nervous system disorders. Armeo^® Power supports specific exercises to increase muscle strength and functional range of motion to improve motor skills. The procedure for using the machine initially involves positioning the patient next to the mechanical arm; if it is not possible to carry out the transfer from the wheelchair to the chair, lifting the armrest on the side of the hemiplegic arm is required. Subsequently, the patient’s arm is inserted into the mechanical one, fixing it with special straps. Finally, the safety measures are set to prevent the patient from exceeding the rehabilitation area and compromising their own safety, that of the operator, or those around them. This machine also provides visual feedback of the work to the patient himself as the rehabilitation focuses on the execution of some games that Armeo^® Power offers. It has been demonstrated through the analysis of various scientific studies how important for the maintenance and recovery of cognitive functions is the application of this feedback to the patient, which allows continuous interaction by the latter throughout the rehabilitation session. In addition to the rehabilitation functions, it also has purely evaluative functions that provide objective and quantitative feedback on the condition of the patient’s limb [21].

It is possible to evaluate the following movements: shoulder flexion–extension; shoulder abduction and adduction; internal and external rotation of the shoulder; elbow flexion–extension; Forearm prono-supination; wrist flexion–extension; and opening and closing of the hand.

The values deriving from the evaluation of the articular range are expressed in degrees; the values deriving from the evaluation of the force are expressed in Newton meters (Nm).

The FMA-UL (FMA) is a performance-based stroke-specific impairment index. It is designed to evaluate motor functioning, sensation, balance, joint range of motion, and joint pain in patients with post-stroke hemiplegia. It is applied clinically and in research to determine disease severity, describe motor recovery, and treatment planning and evaluation.

The scale is composed of five domains and 155 items in total:

-

Motor function in both upper and lower limbs;

-

Feeling;

-

Balance;

-

Range of joint motion;

-

Articular pains.

The motor domain includes elements that evaluate the movement, coordination, and reflex action of the shoulder, elbow, forearm, wrist, hand, hip, knee, and ankle. Scoring is based on the direct observation of performance. Scale items are rated on the ability to accomplish a task using a 3-point ordinal scale where: 0 = cannot perform, 1 = partially performs, and 2 = completely performs.

The points are divided as follows:

-

Motor function score: ranges from 0 (hemiplegia) to 100 points (normal performance), divided into 66 points for the upper extremity and 34 for the lower extremity.

-

Sensation score: ranges from 0 to 24 points, divided into 8 points for light touch and 16 points for positional detection.

-

Balance score: varies from 0 to 14 points, divided into 6 points for sitting and 8 points for standing.

-

Joint range of motion score: ranges from 0 to 44 points.

-

Joint pain score: ranges from 0 to 44 points.

Each of the five FMA domains can be separated to test a specific construct. For example, to assess upper extremity function, subsections specifically addressing upper extremity movement, sensation, joint movement, and pain can be examined without administering the rest of the scale. The FMA score will depend on the number of items included in the subsection selected for testing. In this study, only the values of the domain “motor function for the upper limb” were taken into consideration: Fugl–Meyer upper limb assessment (FMA-UL) [26].

The Stroke Upper Limb Capacity Scale (SULCS) [25] is a simple-to-administer, one-dimensional, hierarchical, and entirely consistent scale that assesses upper limb capacity in post-stroke patients.

It presents 10 items that evaluate the patient’s ability to carry out activities of daily life: item 1—use forearm for support while sitting; item 2—block an object between the chest and upper arm; item 3—slide an object from one side of a table to the other while sitting; item 4—unscrew (partially) a screw cap; item 5—take a glass of water and drink from it; item 6—catch a ball placed in a high corner; item 7—comb your hair; item 8—button up the buttons; item 9—write; item 10—handle coins. Instructions and explanation: The tasks on the list are sorted by difficulty and complexity; they can be performed standing or sitting and must be performed unaided. It is important to rate whether a task can be performed according to the instructions (capable/unable), not the quality with which it is performed; if necessary, it is allowed to repeat the instructions or show the task, and it can be decided whether to start with task 1 or with task 10 by previously assessing the level of motor capacity of the patient’s upper limbs. One should start with task 1 if the level is low, and task 10 if it is high.

2.3. Procedures, Data Collection, and Analysis

All participants were evaluated within a week of entering the facility with SULCS, Fugl–Meyer upper limb assessment (FMA-UL), and the Armeo^® Power. An experienced occupational therapist assessed each participant on the first day using the two outcome measures and a questionnaire to collect demographic characteristics, and the following day proceeded with the evaluation using the Armeo^® Power. The demographic characteristics and results from the outcome measures were inserted into an Excel sheet and the output from the robotic tool was added to analyze the data.

A descriptive and inferential analysis was performed. To evaluate the correlation between SULCS and FMA-UL and strength and joint assessments obtained with the Armeo^® Power, Pearson’s correlation coefficient was used, which reflects the strength and direction of the linear link between variables. The Pearson correlation coefficient can be interpreted as follows: 0 indicates no linear relationship; +1/−1 indicates a perfect linear positive/negative relationship; values between 0 and 0.3 (0 and −0.3) indicate a weak linear positive (negative) relationship through a shaky linear rule; values ranging from 0.3 to 0.7 (−0.3 and −0.7) indicate a moderate positive (negative) linear relationship through a fuzzy–firm linear rule; and values between 0.7 and 1.0 (−0.7 and −1.0) indicate a strongly positive (negative) linear relationship through a firm, linear rule. The significance level was set at a p-value less than or equal to 0.05. All statistical analyses were performed using IBM-SPSS version 23.0.

3. Results

A total of 102 stroke survivors were included in this cross-sectional study, and all participants finished the study. The mean age was 65.83 (13.36), 59.8% of the population was female, and 74.55% of the included population had a plegic dominant hand.

Table 1. Demographic characteristics of the included participants.

Mean (±Standard Deviation)
age	65.83 (13.36)
Number (%)
Gender: female	61 (59.8)
Ischemic	68 (66.66)
Right hand plegic	68 (66.66)
Dominant right hand	85 (83.33)
Plegic dominant hand	76 (74.5)

reports all the demographic characteristics.

The results obtained from the assessments carried out with the SULCS and FMA-UL scales and the joint and strength assessments carried out with Armeo^® Power were correlated using Pearson’s correlation. A correlation was performed between the data obtained from the SULCS and FMA-UL and the joint and strength evaluations carried out with Armeo^® Power.

In

Table 2. Correlation between Stroke Upper Limb Capacity Scale (SULCS) and Armeo^® Power data.

Sulcs	Shoulder						Elbow		Forearm		Wrist		Hand
	Flex.	Exten.	Abd.	Add.	Rot. Int.	Rot. Ext.	Flex.	Exten.	Pron.	Supin.	Flex.	Exten.	Opening	Closing
Item 1 Use the Forearm as A Support to Take Something	0.674 *	−0.775 **	0.599 *	−0.513	0.713 **	−0.156	0.700 *	0.062	0.745 **	−0.381	0.627 *	−0.614 *	0.638 *	−0.244
Item 2 Block Something between the Chest and upper Arm	0.674 *	−0.775 **	0.599 *	−0.513	0.713 **	−0.156	0.700 *	0.062	0.745 **	−0.381	0.627 *	−0.614 *	0.638 *	−0.244
Item 3 Slide An Object across the Table While Sitting	0.674 *	−0.775 **	0.599 *	−0.513	0.713 **	−0.156	0.700 *	0.062	0.745 **	−0.381	0.627 *	−0.614 *	0.638 *	−0.244
Item 4 Unscrew A Cover	0.357	−0.293	0.270	0.044	0.211	−0.188	0.374	0.305	0.295	−0.505	0.353	−0.313	0.427	−0.331
Item 5 Take A Glass of Water and Drink from It	0.674 *	−0.775 **	0.599 *	−0.513	0.713 **	−0.156	0.700 *	0.062	0.745 **	−0.381	0.627 *	−0.614 *	0.638 *	−0.244
Item 6 Catch A Ball Placed in A High Corner	0.522	−0.667 *	0.684 *	−0.278	0.522	0.000	0.570	0.080	0.571	−0.328	0.474	−0.448	0.444	−0.093
Item 7 Comb Your Hair	0.674 *	−0.775 **	0.599 *	−0.513	0.713 **	−0.156	0.700 *	0.062	0.745 **	−0.381	0.627 *	−0.614 *	0.638 *	−0.244
Item 8 Button up the Buttons	0.213	−0.306	−0.127	−0.052	0.061	−0.049	0.229	0.270	0.144	−0.302	0.356	−0.105	0.155	−0.295
Item 9 Writing	0.174	−0.333	0.268	−0.320	0.373	−0.402	0.177	−0.214	0.368	−0.492	0.443	−0.241	0.317	−0.500
Item 10 Handling Coins	0.674 *	−0.775 **	0.599 *	−0.513	0.713 **	−0.156	0.700 *	0.062	0.745 **	−0.381	0.627 *	−0.614 *	0.638 *	−0.244

* p < 0.05; ** p < 0.01.

, the scale items have been correlated with SULCS, and the data analyzed by the joint assessments performed with the Armeo^® Power (range of movement—joint evaluation). As can be seen from the table, there are statistically significant correlations.

SULCS item 1: Using the forearm as a support to hold something has a correlation at p < 0.05 with the results obtained from the joint evaluation performed with the Armeo^® Power of shoulder flexion (0.674), shoulder abduction (0.599), elbow flexion (0.700), wrist flexion (0.627), wrist extension (0.614), and hand opening (0.638). Meanwhile, it has a correlation at p < 0.01 with the results obtained from the joint evaluation carried out with the Armeo^® Power of shoulder extension (−0.775), shoulder internal rotation (0.713), and forearm pronation (0.745).

SULCS scale item 2: Blocking something between the chest and the upper arm has a correlation at p < 0.05 with the results obtained from the joint evaluation carried out with the Armeo^® Power of shoulder flexion (0.674), shoulder abduction (0.599), elbow flexion (0.700), wrist flexion (0.627), wrist extension (0.614), and hand opening (0.638). Meanwhile, it has a correlation at p < 0.01 with the results obtained from the joint evaluation carried out with the Armeo^® Power of shoulder extension (−0.775), shoulder internal rotation (0.713), and forearm pronation (0.745).

Item 3: Sliding an object from one side of the table to the other while sitting on the SULCS scale has a correlation at p < 0.05 with the results obtained from the joint evaluation carried out with the Armeo^® Power of shoulder flexion (0.674), shoulder abduction (0.599), elbow flexion (0.700), wrist flexion (0.627), wrist extension (0.614), and hand opening (0.638). Meanwhile, it has a correlation at p < 0.01 with the results obtained from the joint evaluation carried out with the Armeo^® Power of shoulder extension (−0.775), shoulder internal rotation (0.713), and forearm pronation (0.745).

Item 5: Taking a glass and drinking from it on the SULCS scale has a correlation at p < 0.05 with the results obtained from the joint evaluation carried out with the Armeo^® Power of shoulder flexion (0.674), shoulder abduction (0.599), elbow flexion (0.700), wrist flexion (0.627), wrist extension (0.614), and hand opening (0.638). Meanwhile, it has a correlation at p < 0.01 with the results obtained from the joint evaluation carried out with the Armeo^® Power of shoulder extension (−0.775), shoulder internal rotation (0.713), and forearm pronation (0.745).

Item 6: Catching a ball placed at the top on the SULCS scale has a correlation at p < 0.5 with the results obtained from the joint evaluation carried out with the Armeo^® Power of shoulder extension (−0.667) and shoulder abduction (0.684). Meanwhile, it does not have a p < 0.01 correlation with any of the results obtained from the joint evaluation carried out with the Armeo^® Power.

Item 7: Combing one’s hair on the SULCS scale has a correlation at p < 0.05 with the results obtained from the joint evaluation carried out with the Armeo^® Power of shoulder flexion (0.674), shoulder abduction (0.599), elbow flexion (0.700), wrist flexion (0.627), wrist extension (0.614), and hand opening (0.638). Meanwhile, it has a correlation at p < 0.01 with the results obtained from the joint evaluation carried out with the Armeo^® Power of shoulder extension (−0.775), shoulder internal rotation (0.713), and forearm pronation (0.745).

Item 10: Handling coins on the SULCS scale has a correlation at p < 0.05 with the results obtained from the joint evaluation carried out with the Armeo^® Power of shoulder flexion (0.674), shoulder abduction (0.599), elbow flexion (0.700), wrist flexion (0.627), wrist extension (0.614), and hand opening (0.638). Meanwhile, it has a correlation at p < 0.01 with the results obtained from the joint evaluation carried out with the Armeo^® Power of shoulder extension (−0.775), shoulder internal rotation (0.713), and forearm pronation (0.745).

Unscrewing a lid (item 4), buttoning up the buttons (item 8), and writing (item 9) on the SULCS scale have no correlation with the data obtained from the joint evaluation carried out with the Armeo^® Power, either for p < 0.05 or for p < 0.01.

Unscrewing a lid (item 4), and writing (item 9) on the SULCS scale do not correlate with the data obtained from the joint evaluation carried out with the Armeo^® Power, either for p < 0.05 or p < 0.01.

In

Table 3. Correlation between Stroke Upper Limb Capacity Scale (SULCS) and Armeo^® Power A-force data.

SULCS	Shoulder						Elbow		Forearm		Wrist		Hand
	Flex.	Exten.	Abd.	Add.	Rot. Int.	Rot. Ext.	Flex.	Exten.	Pron.	Supin.	Flex.	Exten.	Closing	Opening
Item 1 Use the Forearm as A Support to Take Something	0.408	−0.509	0.651 *	−0.484	0.461	−0.307	0.608 *	−0.313	0.517	−0.450	0.627 *	−0.337	0.524	−0.227
Item 2 Block Something between the Chest and upper Arm	0.408	−0.509	0.651 *	−0.484	0.461	−0.307	0.608 *	−0.313	0.517	−0.450	0.627 *	−0.337	0.524	−0.227
Item 3 Slide An Object across the Table While Sitting	0.408	−0.509	0.651 *	−0.484	0.461	−0.307	0.608 *	−0.313	0.517	−0.450	0.627 *	−0.337	0.524	−0.227
Item 4 Unscrew A Cover	0.369	−0.318	0.367	0.558	0.241	−0.282	0.283	−0.392	0.360	−0.453	0.412	−0.167	0.441	0.310
Item 5 Take A Glass of Water and Drink from It	0.408	−0.509	0.651 *	−0.484	0.461	−0.307	0.608 *	−0.313	0.517	−0.450	0.627 *	−0.337	0.524	−0.227
Item 6 Catch A Ball Placed in A High Corner	0.527	−0.641 *	0.664 *	−0.543	0.491	−0.403	0.644 *	−0.357	0.418	−0.506	0.692 *	−0.157	0.624 *	−0.166
Item 7 Comb Your Hair	0.408	−0.509	0.651 *	−0.484	0.461	−0.307	0.608 *	−0.313	0.517	−0.450	0.627 *	−0.337	0.524	−0.227
Item 8 Button up the Buttons	0.399	−0.301	0.169	−0.449	0.051	−0.490	−0.146	0.165	0.303	0.102	0.259	−0.004	0.175	0.593 *
Item 9 Writing	0.115	−0.547	0.143	−0.268	−0.116	−0.158	0.271	0.213	0.173	0.053	0.364	0.470	0.260	0.523
Item 10 Handling Coins	0.408	−0.509	0.651 *	−0.484	0.461	−0.307	0.608 *	−0.313	0.517	−0.450	0.627 *	−0.337	0.524	−0.227

* p < 0.05.

, the items of the scale do not correlate with SULCS and the data analyzed by the joint assessments performed with the Armeo^® Power (strength evaluation). As can be seen from the table, there are statistically significant correlations.

Item 1: Using the forearm as a support to hold something on the SULCS scale has a correlation at p < 0.05 with the results obtained by evaluating the strength of shoulder abduction (0.651), elbow flexion (0.608), and wrist flexion (0.627).

SULCS item 2: Blocking something between the chest and upper arm has a correlation at p < 0.05 with the results obtained by assessing the strength of shoulder abduction (0.651), elbow flexion (0.608), and wrist flexion (0.627).

Item 3: Sliding an object from one side of the table to the other while sitting on the SULCS scale has a correlation at p < 0.05 with the results obtained by evaluating the strength of shoulder abduction (0.651), elbow flexion (0.608), and wrist flexion (0.627).

Item 5: Taking a glass and drinking from it on the SULCS scale has a correlation at p < 0.05 with the results obtained by evaluating the strength of shoulder abduction (0.651), elbow flexion (0.608), and wrist flexion (0.627).

Item 6: Catching a ball placed at the top of the SULCS scale has a correlation at p < 0.05 with the results obtained by evaluating the strength of shoulder extension (−0.641), shoulder abduction (0.664), elbow flexion (0.644), wrist flexion (0.627), and hand closure (0.624).

Item 7: Combing hair on the SULCS scale has a correlation at p < 0.05 with the results obtained by evaluating the strength of shoulder abduction (0.651), elbow flexion (0.608), and wrist flexion (0.627).

Item 8: Buttoning buttons on the SULCS scale has a correlation at p < 0.05 with the results obtained by assessing the strength of the hand opening (0.593).

Item 10: Handling coins on the SULCS scale has a correlation at p < 0.05 with the results obtained by evaluating the strength of shoulder abduction (0.651), elbow flexion (0.608), and wrist flexion (0.627).

None of the items on the SULCS scale has a correlation at p < 0.01 with the strength assessments made with the Armeo^® Power.

In

Table 4. Correlation between Fugl–Meyer upper limb assessment (FMA-UL) and A-Roma Armeo^® Power data.

FMA-UL	Shoulder						Elbow		Forearm		Wrist		Hand
	Flex.	Exten.	Abd.	Add.	Rot. Int.	Rot. Ext.	Flex.	Exten.	Pron.	Supin.	Flex.	Exten.	Opening	Closing
Flexor Reflex Activity	−0.091	0.261	−0.295	0.123	−0.143	0.210	−0.113	−0.586 *	−0.257	0.257	−0.008	0.126	−0.165	0.184
Extensor Reflex Activity	−0.091	0.261	−0.295	0.123	−0.143	0.210	−0.113	−0.586 *	−0.257	0.257	−0.008	0.126	−0.165	0.184
Flexor Synergy (Shoulder Retraction)	0.522	−0.500	0.303	−0.295	0.463	−0.097	0.554	0.112	0.596 *	−0.443	0.559	−0.434	0.342	−0.300
Flexor Synergy (Shoulder Elevation)	0.522	−0.500	0.303	−0.295	0.463	−0.097	0.554	0.112	0.596 *	−0.443	0.559	−0.434	0.342	−0.300
Flexor Synergy (Shoulder Abduction)	0.539	−0.581 *	0.472	−0.380	0.501	−0.171	0.578 *	0.016	0.672 *	−0.524	0.580 *	−0.507	0.393	−0.359
Flexor Synergy (External Rotation Shoulder)	0.440	−0.361	0.236	−0.177	0.314	−0.276	0.461	0.159	0.500	−0.622 *	0.502	−0.423	0.389	−0.488
Flexor Synergy (Elbow Flexion)	0.592 *	−0.661 *	0.641 *	−0.315	0.536	−0.274	0.624 *	0.136	0.647 *	−0.605 *	0.572	−0.586 *	0.647 *	−0.399
Flexor Synergy (Forearm Supination)	0.426	−0.510	0.437	−0.366	0.487	−0.542	0.433	−0.033	0.575	−0.754 **	0.599 *	−0.506	0.621 *	−0.681 *
Extension Synergy (Shoulder Adduction)	0.522	−0.500	0.303	−0.295	0.463	−0.097	0.554	0.112	0.596 *	−0.443	0.559	−0.434	0.342	−0.300
Extension Synergy (Elbow Extension)	0.563	−0.665 *	0.571	−0.418	0.544	−0.245	0.609 *	0.229	0.754 **	−0.608 *	0.572	−0.543	0.446	−0.419
Extension Synergy (Forearm Supination)	0.364	−0.475	0.503	−0.248	0.402	−0.428	0.385	−0.015	0.495	−0.702 *	0.520	−0.426	0.518	−0.567
Hand at the Lumbar Level of the Spine	0.392	−0.402	0.440	−0.158	0.280	−0.246	0.430	0.077	0.490	−0.633 *	0.447	−0.410	0.347	−0.435
Shoulder Flex from 0 to 90°	0.455	−0.435	0.388	−0.256	0.351	−0.336	0.482	0.070	0.568	−0.686 *	0.519	−0.486	0.430	−0.532
Shoulder Abduction 0–90°	0.440	−0.361	0.236	−0.177	0.314	−0.276	0.461	0.159	0.500	−0.622 *	0.502	−0.423	0.389	−0.488
0–90° Flexion Shoulder	0.440	−0.361	0.236	−0.177	0.314	−0.276	0.461	0.159	0.500	−0.622 *	0.502	−0.423	0.389	−0.488
Pronation/Supination 0–90°	0.364	−0.475	0.503	−0.248	0.402	−0.428	0.385	−0.015	0.495	−0.702 *	0.520	−0.426	0.518	−0.567
Stability 15° Dorsiflection (Wrist)	0.342	−0.303	0.490	−0.113	0.358	−0.353	0.354	−0.158	0.401	−0.447	0.419	−0.412	0.393	−0.423
Repeated Dorsiflexion (Wrist)	0.321	−0.502	0.443	−0.347	0.404	−0.418	0.359	0.255	0.609 *	−0.536	0.458	−0.375	0.202	−0.574
Repeated Dorsiflexion/Volar Flexion (Wrist)	0.330	−0.395	0.141	−0.284	0.236	−0.229	0.354	0.038	0.430	−0.506	0.464	−0.327	0.240	−0.457
Circumduction (Wrist)	0.322	−0.463	0.427	−0.138	0.161	−0.354	0.346	0.316	0.239	−0.570	0.580 *	−0.303	0.587 *	−0.497
Massive Flex. (Hand)	0.512	−0.420	0.537	−0.314	0.303	−0.017	0.561	−0.118	0.522	−0.569	0.623 *	−0.449	0.453	−0.195
Massive Extension (Hand)	0.522	−0.500	0.589 *	−0.346	0.343	−0.097	0.578 *	0.112	0.596 *	−0.443	0.596 *	−0.475	0.342	−0.256
Hook Grip (Hand)	0.426	−0.510	0.619 *	−0.366	0.365	−0.542	0.457	0.295	0.575	−0.553	0.599 *	−0.506	0.466	−0.635 *
Thumb Abduction (Hand)	0.432	−0.451	0.597 *	−0.352	0.342	−0.508	0.456	0.072	0.524	−0.722 **	0.658 *	−0.505	0.614 *	−0.623 *
Clamp Grip/Opposition (Hand)	0.432	−0.451	0.597 *	−0.352	0.342	−0.508	0.456	0.072	0.524	−0.722 **	0.658 *	−0.505	0.614 *	−0.623 *
Grip of A Cylinder (Hand)	0.426	−0.408	0.619 *	−0.209	0.305	−0.493	0.433	−0.033	0.364	−0.653 *	0.599 *	−0.506	0.776 **	−0.545
Spherical Grip (Hand)	0.366	−0.315	0.550	−0.009	0.198	−0.338	0.380	0.084	0.227	−0.569	0.533	−0.376	0.666 *	−0.413

* p < 0.05; ** p < 0.01.

, the items of the scale have been correlated with FMA-UL and the data analyzed by the joint assessments performed with the Armeo^® Power (range of movement—joint evaluation). As can be seen from the table, there are statistically significant correlations.

The FMA-UL flexor reflex activity item has a correlation at p < 0.05 with the results obtained from the joint evaluations performed with the Armeo^® Power of elbow extension (−0.586).

The extensor reflex activity item of the FMA-UL has a correlation at p < 0.05 with the results obtained from the joint evaluations carried out with the Armeo^® Power of elbow extension (−0.586).

The flexor synergy (shoulder retraction) item of the FMA-UL has a correlation at p < 0.05 with the results obtained from the joint evaluations carried out with the Armeo^® Power of forearm pronation (−0.586).

The FMA-UL flexor synergy (shoulder elevation) item has a correlation at p < 0.05 with the results obtained from the joint evaluations of forearm pronation performed with the Armeo^® Power (0.596).

The FMA-UL flexor synergy (shoulder abduction) item has a correlation at p < 0.05 with the results obtained from the joint evaluations carried out with the Armeo^® Power of elbow flexion (0.582), forearm pronation (0.672), and wrist flexion (0.580).

The flexor synergy item (shoulder external rotation) of the FMA-UL has a correlation at p < 0.05 with the results obtained from the joint evaluations carried out with the Armeo^® Power of forearm supination (−0.622).

The FMA-UL flexor synergy (elbow flexion) item has a correlation at p < 0.05 with the results obtained from joint evaluations carried out with the Armeo^® Power of shoulder extension (−0.661), shoulder abduction (0.641), elbow flexion (0.624), forearm pronation (0.647), forearm supination (−0.605), wrist extension (−0.587), and hand opening (0.647).

The flexor synergy item (forearm supination) of the FMA-UL has a correlation at p < 0.05 with the results obtained from the joint evaluations carried out with the Armeo^® Power of wrist flexion (0.599), hand opening (0.621), and hand closure (−0.681). Meanwhile, it has a correlation at p < 0.01 with the results obtained from joint evaluations carried out with Armeo^® Power of forearm supination (−0.754).

The extensor synergy (shoulder adduction) item of the FMA-UL has a correlation at p < 0.05 with the results obtained from the joint evaluations of forearm pronation performed with the Armeo^® Power (0.596).

The extensor synergy item (elbow extension) of the FMA-UL has a correlation at p < 0.05 with the results obtained from the joint evaluations carried out with the Armeo^® Power of shoulder extension (−0.665), elbow flexion (0.609), and forearm supination (−0.608). Meanwhile, it has a correlation at p < 0.01 with the results obtained from the joint evaluations carried out with the Armeo^® Power of forearm pronation (0.754)

The extensor synergy (forearm pronation) item of the FMA-UL has a correlation at p < 0.05 with the results obtained from joint assessments of forearm supination (−0.702).

The hand at the lumbar spine level of the FMA-UL has a correlation at p < 0.05 with the results obtained from the joint assessments of forearm supination (−0.633).

The shoulder flexion item from 0° to 90° of the FMA-UL has a correlation at p < 0.05 with the results obtained from joint assessments of forearm supination (−0.686).

The shoulder abduction item from 0° to 90° of the FMA-UL has a correlation at p < 0.05 with the results obtained from joint assessments of forearm supination (−0.622).

The shoulder flexion item 0° to 90° of the FMA-UL has a correlation at p < 0.05 with the results obtained from the joint evaluations carried out with the Armeo^® Power of forearm supination (−0.622).

The forearm prono-supination item of the FMA-UL has a correlation at p < 0.05 with the results obtained from the joint assessments of forearm supination performed with the Armeo^® Power (−0.702).

The repetitive dorsiflexion (wrist) item of the FMA-UL has a correlation at p < 0.05 with the results obtained from the joint evaluations carried out with the Armeo^® Power of forearm pronation (0.609).

The circumduction (wrist) item of the FMA-UL has a correlation at p < 0.05 with the results obtained from the joint evaluations carried out with the Armeo^® Power of wrist flexion (0.580) and hand opening (0.587).

The item massive flexion (hand) of the FMA-UL has a correlation at p < 0.05 with the results obtained from the joint evaluations carried out with the Armeo^® Power of wrist flexion (0.623)

The massive extension (hand) item of the FMA-UL has a correlation at p < 0.05 with the results obtained from the joint evaluations carried out with the Armeo^® Power of shoulder abduction (0.589), elbow flexion (0.578), forearm pronation (0.596), and wrist flexion (0.596).

The hook grip (hand) item of the FMA-UL has a correlation at p < 0.05 with the results obtained from the joint evaluations carried out with the Armeo^® Power of shoulder abduction (0.619), wrist flexion (0.599), and hand closure (−0.635)

The thumb (hand) abduction item of the FMA-UL has a correlation at p < 0.05 with the results obtained from the joint evaluations carried out with the Armeo^® Power of shoulder abduction (0.597), wrist flexion (0.658), hand opening (0.614), and hand closure (−0.623).

The thumb (hand) abduction item of the FMA-UL has a correlation at p < 0.01 with the results obtained from the joint assessments performed with the Armeo^® Power of forearm supination (−0.722).

The pincer grip/thumb (hand) opposition item of the FMA-UL has a correlation at p < 0.05 with the results obtained from the joint evaluations carried out with the Armeo^® Power of shoulder abduction (0.597), wrist flexion (0.658), hand opening (0.614), and hand closure (−0.623).

The item gripping a cylinder (hand) of the FMA-UL has a correlation at p < 0.05 with the results obtained from the joint evaluations carried out with the Armeo^® Power of wrist flexion (0.619) supination forearm (−0.633), wrist flexion (0.599), and hand opening (0.776).

The ball grip (hand) item of the FMA-UL has a correlation at p < 0.05 with the results obtained from the joint evaluations carried out with the Armeo^® Power of hand opening (0.666).

The items reflecting the activities of the flexors and the extensors, flexor synergy (shoulder retraction), flexor synergy (shoulder elevation), flexor synergy (shoulder abduction), flexor synergy (external shoulder rotation), flexor synergy (elbow flexion), extensor synergy (shoulder assignment), extending synergy (lumbar pronation), shoulder flexion from 0° to 90°, shoulder abduction from 0° to 90°, downstairs drop 0° to 90°, prone-supination of the forearm, repeated dorsiflexion (wrist), circumduction (wrist), massive flexion (hand), massive extension (hand), hook grip (hand), pliers/opposition of the thumb (hand), taking a cylinder (hand) of the FMA-UL do not have a p < 0.01 correlation with the results of the joint assessments made with the Armeo^® Power. The 15° stability items of dorsiflexion (wrist), repeated dorsiflexion/volar flexion (wrist), and ball socket (hand) of the FMA-UL did not correlate at either p < 0.05 or p < 0.01 with the joint assessments performed with the Armeo^® Power.

In Table 3, the scale items have been correlated with SULCS and the data analyzed by the joint assessments performed with the Armeo^® Power (strength evaluation). As can be seen from the table, there are statistically significant correlations.

The flexor synergy (shoulder retraction) item of the FMA-UL has a correlation at p < 0.05 with the results obtained from the strength assessments carried out with the Armeo^® Power of forearm pronation (0.647) and wrist flexion (0.613). Meanwhile, it has a correlation at p < 0.01 with the results obtained from joint evaluations carried out with the Armeo^® Power of shoulder adduction (−0.710).

The flexor synergy (shoulder elevation) item of the FMA-UL has a correlation at p < 0.05 with the results obtained from the strength assessments carried out with the Armeo^® Power of forearm pronation (0.647) and wrist flexion (0.613). Meanwhile, it has a correlation at p < 0.01 with the results obtained from joint evaluations carried out with the Armeo^® Power of shoulder adduction (−0.710).

The flexor synergy (shoulder abduction) item of the FMA-UL has a correlation at p < 0.05 with the results obtained from the strength assessments carried out with the Armeo^® Power of shoulder adduction (−0.661), forearm pronation (0.669), wrist flexion (0.675), and hand opening (0.644).

The flexor synergy item (shoulder external rotation) of the FMA-UL has a correlation at p < 0.05 with the results obtained from the strength assessments carried out with the Armeo^® Power of shoulder adduction (−0.666) and forearm pronation (0.620).

The flexor synergy (elbow flexion) item of the FMA-UL has a correlation at p < 0.05 with the results obtained from the strength assessments carried out with the Armeo^® Power of shoulder abduction (0.650), elbow flexion (0.650), wrist flexion (0.632), and hand opening (0.648).

The extensor synergy (shoulder adduction) item of the FMA-UL has a correlation at p < 0.05 with the results obtained from the strength assessments carried out with the Armeo^® Power of shoulder extension (−0.620), shoulder abduction (0.602), forearm pronation (0.647), and wrist flexion (0.613). Meanwhile, it correlates at p < 0.01 with the results obtained from the strength assessments carried out with the Armeo^® Power of shoulder adduction (−0.710).

The extensor synergy item (elbow extension) of the FMA-UL correlates at p < 0.05 with the results obtained from the strength assessments carried out with the Armeo^® Power of shoulder extension (−0.651), shoulder adduction (−0.687), forearm pronation (0.646), and hand opening (0.642). Meanwhile, it correlates at p < 0.01 with the results obtained from strength evaluations carried out with the Armeo^® Power of shoulder abduction (0.713) and wrist flexion (0.709).

The extensor synergy item (forearm pronation) of the FMA-UL correlates at p < 0.05 with the results obtained from the strength assessments carried out with the Armeo^® Power of shoulder extension (−0.622).

The hand at lumbar spine item of the FMA-UL correlates at p < 0.05 with results of strength assessments performed with the Armeo^® Power of shoulder extension (−0.637), shoulder abduction (0.578), shoulder adduction (−0.632), wrist flexion (0.594), and hand opening (0.652).

The FMA-UL item shoulder flexion from 0° to 90° correlates at p < 0.05 with the results of strength evaluations carried out with the Armeo^® Power of shoulder extension (−0.584), shoulder adduction (−0.618), forearm pronation (0.636), and hand opening (0.615).

The shoulder abduction item from 0° to 90° of the FMA-UL correlates at p < 0.05 with the results of the strength assessments carried out with the Armeo^® Power of forearm supination (−0.622).

The FMA-UL 0° to 90° shoulder flexion item correlates at p < 0.05 with the results of the strength assessments performed with the Armeo^® Power of forearm supination (−0.622).

The forearm prono-supination item of the FMA-UL correlates at p < 0.05 with the results of the strength assessments carried out with the Armeo^® Power of shoulder adduction (−0.666).

The FMA-UL 15° dorsiflexion (wrist) stability item does not correlate at p < 0.05 with the results obtained from the strength assessments performed with the Armeo^® Power of shoulder extension (−0.660), wrist flexion (0.616), and wrist flexion (0.597).

The 15° dorsiflexion (wrist) stability item of the FMA-UL does not correlate at p < 0.01 with the results of the strength assessments carried out with the Armeo^® Power of the opening of the hand (0.724).

The FMA-UL repeated dorsiflexion (wrist) item correlates at p < 0.05 with the results of the strength assessments carried out with the Armeo^® Power of shoulder flexion (0.595), shoulder abduction (0.594), shoulder external rotation (−0.633), forearm pronation (0.632), wrist flexion (0.696), and hand opening (0.633). Meanwhile, it correlates at p < 0.01 with the results of strength evaluations carried out with the Armeo^® Power of shoulder extension (−0.719) and shoulder adduction (−0.711).

The massive extension (hand) item of the FMA-UL has a correlation at p < 0.05 with the results obtained from the strength assessments performed with the Armeo^® Power of shoulder abduction (0.656), forearm pronation (0.703), wrist flexion (0.613), and hand opening (0.634).

The hook grip (hand) item of the FMA-UL correlates at p < 0.05 with the results of the strength assessments carried out with the Armeo^® Power of shoulder external rotation (−0.650), forearm pronation (0.673), wrist flexion (0.611), and hand opening (0.669)

The items flexor synergy (shoulder abduction), flexor synergy (shoulder external rotation), flexor synergy (elbow flexion), extensor synergy (forearm pronation), hand at lumbar spine level, shoulder flexion from 0° to 90°, shoulder abduction from 0° to 90°, forearm prono-supination), and hook grip (hand) of the FMA-UL do not have a correlation at p < 0.01 with the results obtained from the strength assessment carried out with the Armeo^® Power.

The FMA-UL items flexor reflex, extensor reflex, flexor synergy (forearm supination), repetitive dorsiflexion/flexion, volar (wrist), circumduction (wrist), massive flexion (hand), thumb abduction (hand), pincer grip/thumb opposition, cylinder grip (hand), and ball grip (hand) did not correlate at p < 0.05 or p < 0.01 with the results obtained from the strength assessments performed with the Armeo^® Power.

4. Discussion

The correlation analysis carried out with Pearson’s correlation index shows numerous correlations between the pre-treatment results of the SULCS and FMA-UL scales and the joint and strength evaluations carried out with the Armeo^® Power.

Considering the results of the evaluations obtained with the SULCS and the results obtained with the joint evaluations with the Armeo^® Power (Table 2), some items have a very significant correlation (p < 0.01) with some evaluation movements of the Armeo^® Power, a significant correlation (p < 0.05) with others, and in certain circumstances there is no correlation at all. For example, taking into consideration item 1 (using the forearm as a support to take something), it is possible to see how there is a strong statistically significant correlation between extension and internal rotation of the shoulder and forearm pronation; it has a significant correlation with shoulder flexion and abduction, elbow flexion, wrist flexion–extension, and hand opening. If we think of the movement that the patient is asked to perform, i.e., to place the injured arm parallel to the edge of the table and take an object placed in front of the contralateral arm with the sound limb, the movements just mentioned all form part of the requested item. Flexion–extension and shoulder abduction are involved to bring the arm closer to the table, and elbow flexion allows the forearm to be brought from a position perpendicular to the table to one parallel to it; moreover, the pronation of the forearm is important to allow one to have a firm base on which to lean. Finally, to allow for good stability, both the wrist, in which flexion–extension is required, and the hand, which must be open with the palm facing the table, are important.

For item 2 (blocking something between the chest and the upper arm), it may be seen in Table 2 that in this case too there are numerous interesting correlations to focus on.

In carrying out the item, the patient is asked to place, with the healthy limb, a newspaper under the armpit of the injured limb and to hold it firmly for ten seconds. In this, it is quite clear how flexion–extension, abduction, and internal rotation of the shoulder come into play, movements evaluated with the Armeo^® Power with which the SULCS has both a strong and a weaker significant correlation. The same statistical significance was found for the elbow flexion movements, wrist flexion–extension, forearm pronation, and hand opening. Even if not essential for the item’s success, these latter movements have a significant correlation since the arm must be positioned optimally and, therefore, be along the side, pronated, with freedom of movement at the wrist level.

A correlation at the limits of significance is shoulder adduction, which is very important for the execution of the item.

Item 3 (sliding an object from one side of the table to the other while sitting) also has some interesting correlations to pay attention to. For the item’s success, the patient is asked to move a cloth on the table’s surface. Looking at Table 1 of Section 3, there are various significant correlations, more or less strong, with the movements that are required in the execution of this exercise, such as flexion–extension, abduction and internal rotation of the shoulder, elbow flexion, forearm pronation, wrist flexion–extension, or hand opening. Shoulder flexion–extension, as well as abduction or internal rotation and elbow flexion, allow for a broad movement of the canvas on the table; the flexion–extension of the wrist, on the other hand, participates halfway between allowing movement and the possibility of keeping the cloth firmly under the hand, which in fact must be open. The forearm pronation gives stable support and places the hand above the cloth. Shoulder abduction, with the SULCS, has a correlation that is at the limit of statistical significance since it is very important for the achievement of the item.

Similarly, in item 5 (taking a glass and drinking from it), item 6 (grasping an overhead ball), item 7 (combing hair), and item 10 (handling coins), there are significant correlations between the joint ratings of segments involved in performance and the respective items. For example, in item 5, there is a significant correlation with shoulder flexion–extension, wrist flexion–extension, or hand opening movements, all functional to the gesture of drinking.

Another example can be seen with item 6, in which the patient has to grab a ball placed high up; in this case, a significant correlation can be seen between the item and the joint evaluations of shoulder extension and abduction.

The opposite argument should be made for item 4 (unscrew a lid), item 8 (button up the buttons), and item 9 (write). In these cases, just mentioned, the evaluations of the SULCS are intended to check the fine motor skills of the hand. Here, an important limitation of the Armeo^® Power emerges, since it is unable to evaluate fine motor skills but only the gross motor skills with the evaluation of the opening and closing of the hand. The correlations between these items and all the movements evaluated with the machine were not statistically significant.

The same problem can occur in other correlations, such as that between item 6 (catching a ball) and the opening and closing of the hand. Considering the movement performed by the patient, a significant correlation is expected, but it is not obtained for the problem just mentioned above. The correlations between SULCS and the pre-treatment evaluations of strength carried out with the Armeo^® Power, visible in Table 2 of Section 3, show how the correlations are clearly lower in number and strength than significance. This is because the SULCS aims to evaluate the quality of the gesture and not the force that is impressed in the execution of the movement.

Table 4 and

Table 5. Correlation between Fugl–Meyer upper limb assessment (FMA-UL) and A-Force Armeo^® Power data.

FMA-UL	Shoulder						Elbow		Forearm		Wrist		Hand
	Flex.	Exten.	Abd.	Add.	Rot. Int.	Rot. Est.	Flex.	Exten.	Pron.	Supin.	Flex.	Exten.	Opening	Closing
Flexor Reflex Activity	−0.098	−0.116	−0.122	0.106	0.017	0.008	0.031	−0.023	0.030	0.288	−0.129	0.432	−0.027	0.047
Extensor Reflex Activity	−0.098	−0.116	−0.122	0.106	0.017	0.008	0.031	−0.023	0.030	0.288	−0.129	0.432	−0.027	0.047
Flexor Synergy (Shoulder Retraction)	0.531	−0.620 *	0.602 *	−0.710 **	0.483	−0.410	0.399	−0.410	0.647 *	−0.412	0.613 *	−0.214	0.551	0.191
Flexor Synergy (Shoulder Elevation)	0.531	−0.620 *	0.602 *	−0.710 **	0.483	−0.410	0.399	−0.410	0.647 *	−0.412	0.613 *	−0.214	0.551	0.191
Flexor Synergy (Shoulder Abduction)	0.519	−0.706 *	0.681 *	−0.661 *	0.469	−0.455	0.561	−0.389	0.669 *	−0.462	0.675 *	−0.110	0.644 *	0.109
Flexor Synergy (External Rotation Shoulder)	0.426	−0.507	0.502	−0.666 *	0.320	−0.380	0.349	−0.411	0.620 *	−0.436	0.509	−0.121	0.535	0.382
Flexor Synergy (Elbow Flexion)	0.442	−0.557	0.650 *	−0.570	0.382	−0.389	0.650 *	−0.398	0.527	−0.574	0.652 *	−0.178	0.648 *	0.014
Flexor Synergy (Forearm Supination)	0.225	−0.544	0.383	−0.447	0.041	−0.270	0.474	−0.080	0.419	−0.277	0.503	0.192	0.478	0.437
Extension Synergy (Shoulder Adduction)	0.531	−0.620 *	0.602 *	−0.710 **	0.483	−0.410	0.399	−0.410	0.647 *	−0.412	0.613 *	−0.214	0.551	0.191
Extension Synergy (Elbow Extension)	0.545	−0.651 *	0.713 **	−0.687 *	0.455	−0.450	0.562	−0.374	0.646 *	−0.558	0.709 **	−0.266	0.642 *	0.090
Extension Synergy (Forearm Supination)	0.310	−0.622 *	0.416	−0.489	0.092	−0.331	0.510	−0.125	0.375	−0.326	0.555	0.255	0.544	0.424
Hand at the Lumbar Level of the Spine	0.468	−0.637 *	0.578 *	−0.632 *	0.333	−0.454	0.515	−0.401	0.574	−0.498	0.594 *	0.037	0.652 *	0.291
Shoulder Flex from 0 to 90°	0.414	−0.584 *	0.571	−0.618 *	0.311	−0.419	0.494	−0.389	0.636 *	−0.477	0.564	−0.030	0.615 *	0.300
Shoulder Abduction 0–90°	0.426	−0.507	0.502	−0.666 *	0.320	−0.380	0.349	−0.411	0.620 *	−0.436	0.509	−0.121	0.535	0.382
0–90° Flexion Shoulder	0.426	−0.507	0.502	−0.666 *	0.320	−0.380	0.349	−0.411	0.620 *	−0.436	0.509	−0.121	0.535	0.382
Pronation/Supination 0–90°	0.310	−0.622 *	0.416	−0.489	0.092	−0.331	0.510	−0.125	0.375	−0.326	0.555	0.255	0.544	0.424
Stability 15° Dorsiflection (Wrist)	0.395	−0.660 *	0.525	−0.557	0.343	−0.425	0.616 *	−0.445	0.532	−0.495	0.597 *	0.179	0.724 **	0.218
Repeated Dorsiflexion (Wrist)	0.595 *	−0.719 **	0.594 *	−0.711 **	0.320	−0.633 *	0.379	−0.140	0.632 *	−0.401	0.696 *	−0.062	0.653 *	0.425
Repeated Dorsiflexion/Volar Flexion (Wrist)	0.383	−0.515	0.408	−0.492	0.178	−0.481	0.220	−0.057	0.539	−0.136	0.423	0.061	0.411	0.411
Circumduction (Wrist)	0.170	−0.259	0.155	−0.297	−0.050	−0.442	0.091	0.164	0.207	−0.119	0.342	0.102	0.301	0.508
Massive Flex. (Hand)	0.148	−0.483	0.469	−0.366	0.364	−0.157	0.495	−0.432	0.488	−0.412	0.444	0.021	0.437	−0.037
Massive Extension (Hand)	0.421	−0.569	0.656 *	−0.564	0.558	−0.496	0.483	−0.476	0.703 *	−0.548	0.613 *	−0.231	0.634 *	−0.054
Hook Grip (Hand)	0.393	−0.492	0.549	−0.546	0.332	−0.650 *	0.388	−0.222	0.673 *	−0.509	0.611 *	−0.183	0.669 *	0.250
Thumb Abduction (Hand)	0.119	−0.420	0.370	−0.360	0.128	−0.328	0.415	−0.173	0.476	−0.386	0.460	0.080	0.492	0.296
Clamp Grip/Opposition (Hand)	0.119	−0.420	0.370	−0.360	0.128	−0.328	0.415	−0.173	0.476	−0.386	0.460	0.080	0.492	0.296
Grip of A Cylinder (Hand)	0.000	−0.282	0.272	−0.204	0.041	−0.234	0.474	−0.185	0.306	−0.370	0.359	0.166	0.456	0.187
Spherical Grip (Hand)	0.106	−0.303	0.239	−0.318	0.100	−0.264	0.374	−0.254	0.251	−0.387	0.377	0.150	0.459	0.295

* p < 0.05; ** p < 0.01.

show the correlations between the FMA-UL and the joint assessments and between the FMA-UL and the strength assessments, respectively. Analyzing these two tables, the significant correlations, unlike the correlations with the data of the SULCS scale, are almost the same for the joint and strength evaluations. This happens because the FMA-UL evaluates the quality of the functional gesture and the impressed force in some ways, which the Armeo^® Power can evaluate perfectly.

The items of the FMA-UL are specific for evaluating a body movement; therefore, the significant correlations are reduced to the segment affected by the given movement. The data obtained showed that there is not always congruence between the request for the FMA-UL item and the movement assessed with the Armeo^® Power.

For example, in Table 3, the flexion synergy item (forearm supination) has a strong significant correlation at p < 0.01 with the joint assessment of forearm supination. The same can be said of the item flexor synergy (elbow flexion); in this case, we have a good statistically significant correlation, bordering on being strong, with the joint evaluation of elbow flexion.

These results were not obtained in the items that evaluate hand fine motor and grasping functions, such as the thumb abduction or cylinder gripping items, in which there are no significant correlations due to the same previously mentioned limitation of the Armeo^® Power.

The evaluation capacity of the Armeo^® Power has also been used in other studies [21,28] even if the studies’ objectives were different and the evaluation purposes were not aimed at merely capturing the evaluation capabilities of the Armeo^® Power. No statistical significance was found in the correlations between the data obtained from the evaluations at T1 and T0.

5. Limitations and Conclusions

Robotics is recognized to have great importance in a patient’s rehabilitation process and to provide him with the best possible type of treatment. Through functional robotics for rehabilitation, the therapist’s role is not diminished; on the contrary, it is highlighted even more thanks to the support that technology can provide and will provide in the years to come. The need to correlate the operator-employee evaluations with those carried out with the Armeo^® Power arises because no tools objectively quantify the state of the patient’s limb. In conclusion, from this study, it can be stated that the Armeo^® Power, based on the analysis of the data collected, can be an objective evaluation tool, which can be combined with the operator-employee traditional evaluation techniques, especially when compared to a patient-reported outcome measures (PROMs). Despite the limitations represented by the duration of the evaluation and the times of administration, this multi-evaluation approach has proved to be an ideal solution for the overall evaluation of the post-stroke patient. This is the best solution for including an objective evaluation of the patient accompanied by the PROMs. Although these two aspects have proved to be correlated with each other, an evaluation from both points of view is recommended.