Usability Assessment of Neurostim System: A sEMG Activated Neuroprosthesis Remotely Monitorable and Configurable


 Background: The scientiﬁc community widely explores rehabilitation techniques for individuals with motor disabilities resulting from a stroke since this condition is observed in millions of individuals worldwide.Results: This work presents the usability assessment of Neurostim, a neuromuscular electrical stimulation system triggered by a myoelectrical surface signal, with patients with hemiplegia caused by stroke. The device has a Bluetooth connection with a smartphone application so that conﬁguration parameters can be sent. Usage statistics can be captured and monitored remotely by healthcare professionals, which is the innovation of the presented device following the strong tendency to expand the clinical environment to the patient’s home. The internet can provide healthcare professionals information to perform a detailed follow-up and interact with their patients remotely. Two control platforms were developed for diﬀerent user proﬁles: neuroprosthesis users (Smartphone Application) and health professionals (Web Interface).Conclusions: Neurostim showed to be a robust system application, despite the need for some improvements. It met all the proposal’s expectations, where users were successfully able to operate it, either from a patient or health professional perspective.

Background Globally, stroke is one of the leading causes of death and long-term disability. It consists of a neurological deficit that occurs when there is a lack of adequate blood flow in a specific area of the brain, either by an obstruction (ischemic, 85% of cases) or rupture (hemorrhagic, 15% of cases) of blood vessels. Factors associated with stroke occurrence include individual habits such as inadequate diet, sedentary lifestyle, smoking, alcoholism, aging, hypertension, and diabetes. Consequently, motor disabilities are commonly observed from brain cells' death in areas responsible for this task [1][2][3].
The damage caused by stroke is directly proportional to its etiology, severity, location, patient age, and associated comorbidities. In less aggressive cases, it is common to see problems such as temporary weakness of limb muscles. In more severe situations, a permanent paralysis on one side of the body or even loss of the ability to communicate (expression and interpretation) are typical. In some cases, with appropriate treatment, it is possible to recover from stroke completely.
However, more than two-thirds of the survivors will have some sequela or disability, compromising the quality of life [1][2][3].
Spasticity is one of the most common impairments after stroke. It is recognized as a phenomenon of velocity-dependent increase in tonic stretch reflexes with exaggerated tendon jerks caused by an imbalance of the control signals coming from the Central Nervous System to the muscles. Therefore, there is a muscular hypertonia that prevents voluntary activation for the execution of movements. As an example, the flexor pattern of the elbow, wrist, and fingers is a commonly observed condition in these individuals, that makes it impossible to perform everyday tasks by themselves, such as feeding, combing hair, brushing teeth, typing in a computer's keyboard, etc. [4,5]. Over the years, studies such as [8], [9], and [10] have demonstrated the efficacy of Neuromuscular Electrical Stimulation (NMES) for motor sequelae from a stroke.
The study presented in [11] showed that patients who use NMES trigged by surface electromyography signal (sEMG) could obtain significant improvement in all functional tests, and their cortical activation was shifted from the ipsilateral sensorimotor cortex to the contralateral. This phenomenon could not be noticed in patients who use only NMES, although they also improved in functional tests. In [9], [12], and [13], there are pieces of evidence for the positive effects of NMES initiated by sEMG in various configurations. However, [14] presented a systematic review and meta-analysis of this modality's benefits and concluded that these are larger and evident in chronic cases. In [15,16], the immediate benefits of donning a myoelectric elbow-wrist-hand orthosis in chronic, moderately impaired stroke survivors are investigated. The use of the system significantly reduced upper limb impairment and increased the performance of specific functional tasks.
Given the previous considerations, this work's objective is to evaluate the usability of a neuro-prosthetic system, making it possible not only to use it in places intended for rehabilitation patient's clinic but also at home during the daily life of these individuals. For this purpose, technologies that allow the capture of surface electromyography (sEMG) and Neuromuscular Electrical Stimulation (NMES) were used effectively. The sEMG indicates the user's movement intention, while the NMES helps to complete the movement. The setup of the parameters, activation intensity, the current intensity of the stimulation, and mode of use are customizable for each user through an application installed on a smartphone. Sending patient information to a cloud database allows health professionals to monitor the device's use and send real-time information back to the patient remotely.
1 Results Table 1 contains the results on the pre and post-test questionnaires for patients collected respectively at the beginning and the end of the study. In contrast, Table 2 shows the results for the pre and post-test questionnaires for health professionals collected respectively at the beginning and the end of the study. The scale values used range from 1 to 5, where 1 stands for the "fully disagree" concept, and 5 represents the "fully agree" concept. repetitively [17,18].
NMES was proposed more than 30 years ago. In recent studies, the method has been enriched; it has been combined with a task-oriented approach and trigged through the voluntary effort. In respect of motor relearning, it is vital to ensure personalized treatment, assistance in complex coordinated tasks execution, daily task performance, but with volitional participation of the patient in task execution, avoiding passive mobilization [18,19]. In [15,16], using a myoelectric elbow-wristhand orthosis significantly reduced upper limb impairment and increased specific functional tasks' performance.
Neurostim is a neuroprosthetic device which sEMG trigger the electrical stimulus application that can be used both, in the therapeutic and domestic environment, under the supervision and according to the guidelines and parameters previously provided by the health professional, who can remotely supervise its use, being what sets it apart from the others. Thus, a patient can perform therapy more frequently and also use it during daily task performance.
Overall, the first impression of the patients, shown by the pre-test questionnaire (Table 1) given by the Occupational Therapist responsible for the session so that the patient could use the application.
As for the complexity of using the system, expressed by patient P1's scores in questions 3 and 8, it is believed that, as he was the only one who had this opinion and was the first to be presented to the study and to attend the sessions, it could have been a consequence of the inaugural session and the learning curve of the technical aspects of the presentation of the device, which was an activity performed by the For question 2, there is an improvement of 1-point for patient P2, converging to the same opinion of patients P3 and P4, and a 2-point worsening for patient P1, which reaffirms his initial perception about finding the system complicated.
When questioned, the patient states that there was technical implications, but did not know what aspects he referred to, preventing him from using the system. This feeling is in agreement and also justifies the scores to questions 4 and 10. For question 6, a 2-point worsening is noted for patient P3. When questioned, this impression was because the device stopped working at the beginning of the ninth session, which had to be interrupted and subsequently canceled, so that the repair could be performed. Undoubtedly, this type of event directly reflects the reliability that the patient has on the device.
For the Occupational Therapists, it is evident from Although the evaluation was carried out in a controlled environment, where qualified professionals performed the placement of Neurostim, and with a small sample of users, all of them could interact with the system brought evidence of some potential improvements before allowing system application in unmonitored situations: Adjustment of the bracelets that fix the electrodes, integrating them in one peace, allowing the user to wear it without assistance.
Increased drawing area on the Smartphone screen for dynamic objects, making it easier to interact with buttons, sliders, etc.
Confirmation of the end of the procedure before sending data to the server prevents unexpected stimulation during transmission.
Adjustment in the scales of the axis values obtained by the sEMG and values chosen for the sensitivity parameter.
Confirmation request to end procedure to avoid finishing by an accidental touch on the screen.

Conlcusion
As the main contribution of this work, the literature review showed that a combined initiative between NMES and sEMG applied to a portable device, with the ability to send and receive data connecting to a smartphone application, and providing support tools and remote monitoring for health professionals, had not yet been proposed. The pioneering of this development will undoubtedly stimulate continuity in improving the design and development of related tools. Given the benefits that an integrated solution for low-cost rehabilitation can bring, it is expected that this platform will become versatile enough to make it a continuous use item, outside of rehabilitation activities, being incorporated into the patient's day-to-day life until its use is expendable.
Neurostim showed to be a reliable system application, despite the need for some improvements. It met all the proposal's expectations, where users were successfully able to operate it, either from a patient or health professional perspective. Its design was versatile enough for the user to feel comfortable wearing it as a physiotherapeutic element in the rehabilitation center, with intentions of expanding the use to a practical way outside the clinic to help perform daily tasks that are no longer possible due to muscle disability. The smartphone application was designed so that a simple interface allows the user to perform all requested activities with ease, preventing eventual errors during the stimulation procedures and avoiding behaviors that would not be expected. The web interface was also developed so the health professionals can have relevant information about patient's usage and perform detailed analysis to perform adjustments to maximize effectivity to patients' treatment, all remotely.

Neurostim
Neurostim usage aims to rehabilitate the upper limbs (hands and wrists) of patients with hemiplegia. It is an entirely custom made neuroprosthesis, which relies on the application of Neuromuscular Electrical Stimulation (NMES) activated by surface myoelectric signal (sEMG). NMES and sEMG technologies are extensively known and used in clinical rehabilitation programs. Figure 1 illustrates the device mounted and being used by a user. The innovation corresponds to the proposal of daily use in all environments (not only at the rehabilitation center). For this prospect, two control platforms were developed: one for operating the neuroprosthesis and another platform for remotely monitoring its activities, allowing the health professional to consume data from the device, and to set up parameters of usage, which includes physiotherapy agenda, flow monitoring, and qualitative analysis.

User Interface
The user interface is an Android smartphone application, built using Android Studio The application can also act as a bi-directional communication channel between the healthcare professional and the patient, so setups, adjustments of the usage parameters, and information sending about the usage can be made remotely through the internet (either with a Wi-Fi connection or cellular data).

It is meant to be run in
Users of this application present motor function disabilities in one side of their body, commonly causing them to lose the wrist and fingers' movements partially.
The interface was developed considering the placement of buttons and dynamic objects on the screen to be easily reachable while using just one hand.
All data is stored locally, considering that the smartphone might not have an active internet connection by the time the procedure ends; thus, synchronization cannot happen. With this mechanism, synchronization can be trigger later on whenever a stable connection is available. Figure 2 shows one of the application screens that contain the live usage data from a user, where the user can also control the intensity of the electrical stimulation and the sEMG sensibility.

Health Professional Interface
The web interface for the health professional was developed in HTML5, PHP, and JavaScript. This interface is a two-way communication channel with the neuroprosthesis, providing the necessary information to configure its use and an instrument of interaction between the patient data and the health professional. By synchronizing the captured neuroprosthesis data from the respective user's smartphones, the healthcare professional can fetch the device's usage history information, so progression analysis from a clinical standpoint can be done and remotely perform the necessary adjustments, as all the data will be stored in a cloud MySQL database.
The web interface can display a significant amount of information in a graphical format, simplifying the data's understanding and providing detailed information as needed for the health professional. It also has tools to automatically indicate to the health professional if the scheduled session was executed following the previously set parameters. Hence, it becomes easier to evaluate the progress and forecast the upcoming sessions of each patient. Figure 3 shows one of the application screens that contain the usage data from a patient, received right after the user completed the session.

User profile
The clinical usability trials were performed with patients over 18 years old, of both genders, with clinical and radiological diagnosis of stroke, motor disabilities in an upper limb due to the presence of right or left hemiparesis, who already be in a rehabilitation program using functional electrical stimulation in muscle groups for wrist and fingers extension and who already use smartphone as a daily routine activity.
All the participants were granted in writing of informed consent to participate in the study. It is important to note that not only the patients but also the Occupational Therapists (referred to as Health Professionals) were part of this study, showing the facilities and difficulties found during the assembly of the neuroprosthesis and the use of the respective interfaces intended for them.
The data collection was done with four patients in the rehabilitation process, all with right spastic hemiplegia, two were male, and two were female, ranging from 19 to 46 years old. Initially, the objective was to carry out the study with at least ten patients. However, the inclusion criteria did not allow such amount in the period available to carry out the study and to provide proper documentation with the results. The health professional group had two occupational therapists responsible for the selected patients' rehabilitation program's regular activities, both female being 35 and 56 years old.

Evaluation sessions
The procedures' objective was to verify whether the developed interfaces are in- The patients were submitted to 10 sessions, these being once a week, lasting one hour each. The session's duration considers the placement of the neuroprosthesis by the therapist, execution of activities requested to both therapist and patient, disassembly of the neuroprosthesis, and oral evaluation of the group session. Table 3 lists the activities requested for each participant. In all sessions, patients were recorded using the application that controls the neuroprosthesis. In this way, one could monitor what facilities and difficulties each one had for the given tasks. In the case of Occupational Therapists, the use of the web interface was monitored by recording directly from the operating system's graphical environment using Webex Recorder (Cisco) software. Data were also collected to identify the eventual difficulties of the users while operating the system.