Clinical Interpretation of Working Volume and Weight Support in Upper Limb Robotic Neurorehabilitation after Stroke

Round 1

Reviewer 1 Report

Thank you for the authors for the hard work on this manuscript and letting me read it. Unfortunately, I am not recommending acceptance of the manuscript at this point.

My problem is basically the focus of your research question. I think the investigation of the clinical implication of robotic parameters is extremely important and needs more information. However, I do not believe you selected appropriate parameters for both the clinical side and the robotic side to support the vision you are presenting.

Let me expand more in detail:

Coming more from the clinical side of research I have to say, the thing that you are missing is the great heterogeneity in the clinical presentation of stroke. I very much disagree with the premise that “just” two parameters, especially those two parameters will allow anyone to capture clinical status in stroke and effectively personalize interventions for stroke survivors. You will need more parameters.

I think you present the problem with your premise correctly in several parts of your manuscript, for example in the Abstract line 26: “These results suggest caution in using robotic parameters as outcome measures because they follow the general improvement of the patient but complex relationships with clinical features are possible”. I completely agree with this but this should have been something already evident at the design stage of your study which should have convinced you to expand your set of parameters.

The second part of the problem is your clinical measures that you use for comparison, they are very general clinical measures measuring a very general form of impairment. You need more specific clinical parameters focused on specific aspects of upper extremity motor impairment to evaluate these two parameters. You have on such parameter already in there (MAS) but spasticity is just one possible aspect on impairment in stroke, you need other more specific parameters like that to evaluate the presentation of upper extremity motor impairement.

You have something extremality valuable in your study data, that you just have gravely misused so far in my opinion and that can give you exactly that specific information. Somethings that can assess a lot of the different aspects of movement, from reaction, to coordination, to accuracy and more is motion capture. Your kinematic and kinetic datasets have the potential to provide you very specific to distinguish different aspects and target them during the robotic intervention. You just use them for range of motion which you already have from your workspace but they can give you so much more.

Also really think on what parameters might be best on the robotics side. If you pick just two parameters I would hypothesis that workspace and weight support are unlikely to ever provide a full pictures just based on their clinical relevance. Clinical intervention research has been more studied so far than robotics. While clinical assessments obviously have clear limitations of what they can measure they provide you information on what should measures of movements are worth assessing and targeting with intervention. Yes, someone with less work space and more weight support will be more impaired. But if you look at capturing the impairment of a single person with stroke just assessing their maximum range of motion and their need of support is very limited in its ability to capture a lot of the interindividual differences of stroke survivors that are generally observed in clinical practice. Try to learn from your clinicians. I do not think that they will only assess the maximum range of motion and based on that decide on their set of interventions. What are the 5-6 aspects of impairment they first assess and what is that outcome measure and how does this outcome measure translate to robotics. Try to stay away from just very general measures of being “more impaired” or “less impaired”.

I think you present in your discussion what actually should have been done and how I think your study would become very interesting and valuable. Page 7 line 262: “It could also allow to define the best setup in terms of kinematic and kinetic robotic parameters for each clinical condition, for example defining when is better to train the patient on a large volume with a high support and when is it better to have a small volume with low support.”

For the future vision of using parameters from robotics to personalize intervention, you need to investigate a considerably larger set of potential candidates that capture more specific aspects of impairment. Your Kinematics and Kinetics should be able to provide you with a very clear picture of the individual impairment. This then will allow you to optimize any and all robotic parameters to just that.

I am sorry but I hope at least in some form I was helpful to you for you next submission.

The Reviewer.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

I received the paper after its first revision and I will begin by stating that the paper is very good, easy to read and follow and provides important data regarding a more and more challenging task for the future. Based on the current statistical data regarding the age span of the population and its associated diseases, the management of patients with motor deficit due to different neurological affections will depend more on technology in the future (due to increased lack of personnel) and it is critical to understand and correlate the work of complementary teams in order to provide a comrehensive solution where technology can help people recover faster and better targetting ultimately an increased quality of life. 

I will point out some suggestions for the paper:

1. Please try to avoid the use of the term neurorobotic as neurorobotics is, based on its definition, a totally different thing as robotic assisted rehabilitation (even though this is mainly caused by different neurologic pathologies).
2. In the abstract (line 19) when you refer to the spatial volume maybe it is better to specify that is the volume covered by the upper limb even though you measure it at the patient's hand level. 
3. Line 82: bioengineeristic is understandable as meaning but it is not a real word;
4. In the introduction you mentioned both at the level of the upper and the lower limb the importance of the correlation of the robot parameters (active joints movements) and the patient joints. A theoretical model transfered than in the control of the robot which proposes a particular kinematic modeling of the robot to achieve directly that is ilustrated in Husty, M. et al, An algebraic parameterization approach for parallel robots analysis, Mechanism and Machine Theory 140, 245-257.
5. Line 93, the term "deserved" is not properly used there.
6. I would kindly request you to define your motivation for using patients with chronic stroke (34 months average) as some medical studies illustrate that rehabilitation results 2 years after the stroke are much less impressive than before. To make myself clear, this is NOT a critic to the study, we need to better understand how we can help all patients but a motivation for the selection of these patients would be nice. Also, have you considered the assessment of other chronical diseases which also could benefit from such therapy? In: Major, Z.Z.; et al Comparative Assessment of Robotic versus Classical Physical Therapy Using Muscle Strength and Ranges of Motion Testing in Neurological Diseases. J. Pers. Med. 2021, 11, 953., the potential positive imapct of robotic therapy was demonstrated with a non-comercial device (so less performant than Armeo Power) on the physical rehabilitation of patients with different types of neurological diseases. 
7. Could you add maybe (I read the paper twice but maybe I missed it) a list with the exercises you carried out with the patients?
8. This is just a question: have you carried out similar exercises with the same type of patients using manual rehabilitation and if yes did you obtain better results using the robot?
9. For the weight support which is the measurement unit? Percentage of the total upper limb weight? (tables 1 and 2)
10. As this is an online publication, could you use colors to better differentiate the data in figure 2? I suspect that each dot represent the performance of a patient...or?
11. On line 148 the term "administered" whern referring to the scales it is not the best one, in my opinion.

On the overall I enjoyed the paper I appreciate the quality of the work and I totally support the conclusions, especially the statement about the interdisciplinary collaboration which would lead, in the end, to better life quality for the patients. 

I would like to congratulate the authors and I consider that the paper can be published with minor revisions. 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

The authors used the objective measurements of kinematic and kinetic parameters (the spatial volume reached by the patient’s hand and the weight support provided by the robot) and six clinical scales to evaluate the efficacy of upper limb robotic rehabilitation in patients “with stroke” (Abstract). The title of the paper includes the phrase „after stroke” which seems to be a good idea, instead „with stroke” used throughout the manuscript, because patients examined in this paper represented the chronic phase of this disease. Unification of the stroke description is necessary throughout the text.

The number of 40 patients enrolled on this study is good enough to conclude about the effects of rehabilitation with  Armeo®  Power device, but in my opinion, two weeks (with ten sessions of treatment) is a short period to provide a clear conclusion about its stability. This issue should be briefly “Discussed” because of its importance.

Clinical scales to evaluate the upper limb mobility, strength, spasticity, pain, neurological deficits and patient’s independence (it would be nice to include additionally the word “patient’s independence” in lines 21-22), are appropriate tools for descriptions the patients’ health status before and after the therapy.

Before the treatment authors found correlations between the working volume and spasticity, as well as the weight support with upper limb strength, pain, spasticity and neurological deficits. Two weeks of robotic rehabilitation brought improvement in all the clinical scores as well as the two parameters improvement (line 24 - which parameters? the introduction of words “kinematic and kinetic” would be proper). It’s weird that the percentage of changes of kinematic parameters did not correlate with changes in clinical scores (needs wider possible explanation in the Discussion section, it was only pointed as the limitation).

The authors concluded about the low utility of robotic parameters without a comparison of clinical scores as outcome measures of robotic rehabilitation progression. According to the description in Introduction section and the results of this study it may be due to the short duration of applied rehabilitation, which should be preliminary included in the study design.

M&M and Results sections are clearly and concisely presented.  

Conclusions are in agreement with statement of Italian Consensus Conference on Robotics in Neurorehabilitation.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

The authors present a very interesting topic on the use of robotic therapy used in post-stroke rehabilitation, and devices types of assessments compared to clinical assessment in post-stroke patients.

I have a few minor suggestions:

The authors should provide the power of the sample size and the effect size.

The authors should provide descriptive data regarding the first clinical assessment, to identify if the patients were included in mild or moderate post-stroke status.

The authors should also emphasize in the introduction, but also in discussion and further in limitation ( as a short time of therapy- 2 weeks), the neuroplasticity phenomenon which can occur also in chronic stages of post-stroke, linked with the visual and proprioceptive feedback provided by the Armeo Power device ( front screen, and exergaming, plus upper limb kinematic chain motion felt by the patient).

Author Response

Please see the attachment.

Author Response File: Author Response.pdf