Electromyography data recorded from orbicularis oculi muscle during a novel grasping pilot study

Abstract

Developing assistive devices for quadriplegic people is very difficult because there are only a few ways in which we can take control commands from the patients. The challenge is to fulfill all the functional and non-functional requirements of the patients using those limited control commands. It is reported in the published literature that the abandonment rate of already available assistive devices is very high. We require more and more research in this domain so that we can develop more effective devices and control algorithms. One hindrance in the development of these devices and control algorithms is the limited access to machines to receive required data from the patients to control assistive devices. In this research, we have proposed a novel way to control assistive devices and presented data for the development of the algorithm too. We recorded the data from the orbicularis oculi muscle from 5 healthy volunteers. Volunteers included 3 males and 2 females, aged 29.5 ± 2.1 years. The recording was done with the help of the RMS Salus 2c EMG machine.

Introduction

Quadriplegia is a condition in which a person cannot move his arms and legs. The general cause of this condition is an injury to the spinal cord. Published literature reports an average of 500,000 fresh cases of quadriplegia annually (Mathur et al. 2015).

Although the situation of spinal cord injuries is terrible for all the patients, it is worse for the patients of poor countries. Most patients suffering from quadriplegia in poor countries are industrial laborers having an average family monthly income of $15 or less. Published survey report’s that about 50% of the patients suffering from this condition are from nuclear families (Mathur et al. 2015; Bian et al. 2016; O’Bard et al. 2017). Another important takeaway from these surveys is that most of the patients, i.e. (80–90%) are male. Hence, in many cases, the caregiver is a woman. Keeping the fact in mind that a quadriplegic patient depends on caregivers for all of its needs, it becomes very difficult for a caregiver to attend to all the needs of the patient. An assistive device/self-help device that can address some of the functional and non-functional requirements of these patients can make the task of caregiver easy.

It has been reported in the literature that the abandonment rate of already available assistive devices for quadriplegic people is very high. Major reasons for the abandonment are the cost of the device, the device not addressing the actual issues, and the patient’s inability to accept the technology (Jiang et al. 2018; Jafar and Nagesh 2020). It has been reported that there is a vast gap between the requirements of the quadriplegic subject and the solutions available for them in the published literature (Ruíz-Serrano et al. 2013; Chowdhury et al. 2018; Jiang et al. 2018; Jafar and Nagesh 2020).

We conducted a literature review to find the current state-of-the-art in the development of assistive devices for quadriplegic people. We did a search using keywords (‘Self-help devices’ OR ‘Assistive Devices’ OR ‘Assistive Product’ OR ‘Assistive Technology’) AND ‘Quadriplegia’ in Science Direct, Pubmed, IEEE Xplore digital library, and Web of Science databases. Original articles, technical and case studies, conference articles, and literature reviews published between 2014and 2019 were used for the study. The research group published the detailed review in the cited article (Jafar and Nagesh 2021). From the review of the literature, we can categorize Assistive devices which work on physiology-based signals in the following categories.

Electroencephalography (EEG) Based devices: EEG-based Self-help devices (SHD) use signals generated in the brain. To use these signals, to develop SHD, we need to classify them efficiently. But the classification of these signals is very difficult due to low spatial resolution (AL-Rousan and Assaleh 2011; Khalid et al. 2015). There is a lot of noise present in these signals too, which makes it difficult to classify them. We can increase the classification accuracy by placing electrodes strategically in specific locations. This will help us in extracting multiple signals simultaneously and using a combination of patterns to identify a specific task (Squires et al. 2013; Schmalfuß et al. 2016). But placing these electrodes in these strategic locations is a challenge, as a minor change in position can have a tremendous impact on classification algorithms.

Voice recognition-based SHD: These kinds of devices use voice-based instructions to receive commands. We use these devices in smart homes to control electronic devices (Laumann et al. 2015; Jiang et al. 2018; Jafar and Nagesh 2021), Voice-controlled smart wheelchairs (Nam et al. 2014; Schmalfuß et al. 2016; Jafar and Nagesh 2021). The challenges with these devices are that the environmental conditions significantly affect classification accuracy. Misinterpreted commands can lead to sudden changes in velocity, the direction of the wheelchair, which can cause injury or frustration in these people (Hakonen et al. 2015; Jiang et al. 2018).

Tongue movement-based devices: These devices track the movement of the tongue and, based on these movements, SHD is controlled. They use external devices placed inside the mouth, which causes lots of speech and hygiene-related problems (Linskell and Bouamrane 2012). It has been reported that implantable tongue motion tracking devices can cause infections and other complications in subjects (26).

Electromyography (EMG) based devices: They are the devices that record the electrical signals generated when a muscle contracts. They have a better signal-to-noise (SNR) ratio as compared to the other type of physiological signals. They require less user attention as compared to the other type of signals (Nam et al. 2014;  Hakonen et al. 2015). They are less affected by environmental noise. Although these signals have lots of advantages, they have some disadvantages too. All quadriplegic persons can’t use them, as their application is limited to the physical condition of the subject. It is a challenge to identify multiple tasks while single commands can be identified easily.

We can conclude from the review that EMG-based devices are the best choice for developing self-help devices if the number of commands required to operate SHD is less. As these devices require voluntary control of muscles, for quadriplegic people only a limited amount of muscles can be used to generate the control signal as they cannot move their body below their neck. One muscle that is the best candidate to be used to control SHD by these people is the orbicularis oculi muscle.

In this paper, we have presented the recorded EMG signals which can operate an Assistive Device. With the help of this dataset, researchers can develop algorithms to operate assistive devices which can be operated by voluntary eye blinks. The researchers can also use this data to improve the currently available algorithms, which can address the unmet function and non-functional requirements of the users in a more efficient way.

Theory

The motor components of our nervous system control and execute the motor functions (movements) of our body. We have two types of motor components in our body.

• Visceral Motor System

• Somatic motor system

The Visceral motor system handles the involuntary control of muscles, while the somatic motor system handles voluntary control muscles. Hence, if a user wants to control a self-help device with the help of the movement of muscles, then the somatic motor system is the best candidate for that. We showed the motor components in Figure 1.

Figure 1. Motor components.

In this study, we aim to get control signals from a quadriplegic person. As a quadriplegic person in most cases can only move their body above the neck, we chose EMG signals generated during eye-blinks to receive control commands from the patient as they are less affected by environmental noise, they have a better signal-to-noise ratio and require less user attention.

Mechanism of Eyeblink: The eye blink occurs when the eyelid opens and closes. The following muscles control the movement of the eyelid.

• Levator palpebrae superioris muscle

• Muller’s muscle

• Frontalis muscle

• Orbicularis oculi muscle

For the opening of the eyelid, the Levator palpebrae superioris muscle is majorly responsible, with minor contributions from the muller’s muscle and frontalis muscle, while for the closing of eyelid orbicularis oculi muscle is responsible.

Orbicularis oculi muscle has two major parts:

• Orbital part

• Palpebral part

The orbital part of the orbicularis oculi muscle is responsible for the voluntary closure of the eyelid while the palpebral part of the orbicularis oculi muscle does the involuntary closure. In this study, we have placed electrodes on the orbital part of the orbicularis oculi muscle to record EMG during eye voluntary closure of the eyelid.

Methods

Inclusion and exclusion criterion

Inclusion

1. Volunteers with ability to understand the research requirements, signed informed consent form and consent to any restrictions applicable during the study.

2. The age of volunteers was above 18.

3. Volunteers who were completely healthy based on a medical assessment.

Exclusion

1. Volunteers with visual impairment

2. Volunteers with hearing Impairment

Experiment scenario

Following strategy was followed to perform the grasping operation

1. The subjects were asked to view a computer screen where images of different hand configurations were flashed in a loop.

2. Subjects were asked to blink voluntarily whenever the image of their choice was flashed.

3. Whenever a blink was detected, an audio message was played asking if you wanted to do the selected grasp.

4. The subject was again asked to confirm by blinking, if blink was detected wrongly then the subject was asked to do nothing.

5. If the subject confirms the grasp selection then commands were sent to the actuator.

6. If no grasp was detected till the end of the loop then the loop was re-initiated (Figure 2)

Figure 2. Grasp selection experiment.

Recording of data

In this study, we have placed electrodes on the orbital part of the orbicularis oculi muscle to record EMG during voluntary closure of the eyelid.

We recorded surface EMG from the orbital part of the orbicularis oculi muscle of 5 healthy subjects. Subjects included 3 males and 2 females, age 29.5 ± 2.1 years. We recorded each subject 2 times a day for two days. The recording period was chosen randomly from 9 am to 5:30 pm. Each session was not less than 4 h apart.

The recording was done with the help of the RMS Salus 2C EMG machine, provided by RECORDERS & MEDICARE SYSTEMS PVT. LTD., India. Recording software was RMS EMG Salus V.7.7.1. The study was approved by the Ethical Committee of the Delhi Technological University, Delhi and all participants provided informed consent before their participation. We have shown the details of the experiment in Figure 3.

Figure 3. Signal recording trial timeline.

Results

In each recording session, 4 trials were recorded. Each day, recording was conducted in 2 sessions. As we did the recording for 2 days, hence total of 16 trials were recorded. We have shown the results of the experiment in Table 1, where yes represents a trial where grasp was done successfully, while no represents the trial where grasp was not performed successfully.

Table 1. Results of grasping experiment.

S. No	Participant 1	Participant 2	Participant 3	Participant 4	Participant 5
Trial 1	Yes	No	Yes	No	No
Trial 2	Yes	No	No	No	Yes
Trial 3	Yes	No	No	No	Yes
Trial 4	Yes	No	Yes	Yes	No
Trial 5	Yes	Yes	Yes	Yes	Yes
Trial 6	Yes	No	Yes	No	Yes
Trial 7	No	Yes	Yes	No	Yes
Trial 8	No	Yes	No	Yes	Yes
Trial 9	Yes	Yes	Yes	No	Yes
Trial 10	Yes	Yes	Yes	Yes	Yes
Trial 11	Yes	Yes	Yes	Yes	Yes
Trial 12	Yes	Yes	Yes	Yes	No
Trial 13	Yes	Yes	Yes	Yes	Yes
Trial 14	Yes	Yes	No	Yes	Yes
Trial 15	Yes	Yes	Yes	Yes	Yes
Trial 16	No	Yes	Yes	Yes	Yes

Author contribution

Jafar, Mohd Rizwan contributed to conception and design of the research, acquisition of data, analysis and interpretation of data, drafting the manuscript, revision of the manuscript for important intellectual content. All authors read and approved the final manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

‘The data that support the findings of this study are openly available in Mendeley Data [Jafar, Mohd Rizwan. 2022. EMG Data from Eyeblinks. Mendeley Data, V2. DOI:10.17632/nm535v73jj.2] at https://data.mendeley.com/datasets/nm535v73jj/2.