An Updated Review on the Impact of an Eye-Controlled Wheelchair on Quality of Life of Patients with Neurological Disorders and theirCaregivers

2.1.1. Participants’ Information

Subject A is a 64-year-old female patient with an endomorphic (obese) build who has been residing in the home-care centre for the past five years. Her past medical history includes right breast cancer, for which she underwent a mastectomy in 2017. Additionally, she has been living with diabetes mellitus, hypertension, and dyslipidaemia for the past 10 years and is on medication. Initially, she was able to walk independently; however, three years ago, she suffered a fall while bathing, resulting in an injury to her lumbar region and a left femoral neck fracture.

Her recent spinal cord injury is linked to a past injury at the lumbar region (L4) sustained during a school hockey match, which led to a hairline fracture. As a result, she developed an L1 spinal cord injury, leading to paraplegia and confinement to a manual wheelchair for mobility. As she ages and her health declines, she has experienced significant upper limb muscle weakness. Currently, she relies on both her manual wheelchair and a caregiver to assist her in moving from her room to other locations within the facility, depending on her needs and preferences.

Examples of scenarios where Subject A requires mobility assistance include:

• Attending spiritual talks organized by the home-care center at the Surau within the compound.
• Meeting and chatting with friends from different clusters or dorms.
• Taking a walk after prolonged periods of lying down or watching television in her room.
• Visiting the in-house clinic to consult with nurses and collect medications.

(Note: The Surau is an Islamic worship place similar to a mosque but with a smaller structure.)

Subject A further expressed that, at times, she chooses to remain in bed rather than attend events at the Surau or socialize with her friends due to the unavailability of caregivers, who are often occupied with assisting other residents.

Her primary concern can be summarized as participation limitation, leading to a reduced QoL— specifically in terms of independence and social interaction. Despite her health challenges, she is sociable, enjoys engaging in conversation, and has a good sense of humour.

Subject B is a 75-year-old female patient with an endomorphic (obese) build who has been residing in the home-care centre for the past seven years. She was diagnosed with atypical T-cell lymphocytes in 2013 and completed radiotherapy. Since 2009, she has also been living with diabetes mellitus and hypertension, for which she is on medication.

Due to her age and weight, she developed bilateral knee osteoarthritis (Stage 2) in 2010. In 2011, she suffered an ischemic stroke, which resulted in paraplegia. Additionally, aging and declining health have led to poor upper limb muscle strength, further limiting her mobility. Since then, she has been dependent on a manual wheelchair.

Despite her condition, Subject B remains cheerful and enjoys eating, talking, and meeting people. Her hobbies include attending talks or gatherings organized by the home-care center and socializing with friends. She currently relies on both her manual wheelchair and a caregiver for mobility assistance.

Examples of scenarios where Subject B requires mobility assistance include:

1. Attending talks organized by the home-care centre.
2. Meeting and chatting with friends from different dorms.
3. Taking a walk after socializing with her roommates.
4. Visiting the in-house clinic to consult with nurses and collect medications.

She further mentioned that, at times, she remains in her room rather than attending events at the main hall or visiting her friends' rooms due to caregivers being occupied with assisting other residents (e.g., bathing or feeding them). Unlike Subject A, she does not engage in specific hobbies but enjoys chatting with friends by the window and watching television, regardless of the program being aired.

Similar to Subject A, Subject B’s primary concern is participation limitation, leading to a reduced QoL—particularly in terms of independence and social interaction.

2.3.1. Usage of the Wheelchair

The usage of the eye-controlled wheelchair was measured based on the number of sessions completed, session duration, and distance travelled. The distance was calculated from the wheelchair’s start to the endpoint using a BOSCH GLM 40 laser range finder, with both points marked by neon-coloured stickers. A designated box at the start point served as a laser barrier to ensure consistent measurement. The researcher, standing at the endpoint, aimed the laser at the box, and the displayed value was recorded in Google Sheets. Stickers were removed after each measurement to prevent errors, with distances measured at both the start and end of each session.

Session duration was recorded using a smartphone stopwatch from the start to the end of each session, with the results saved directly in Google Sheets. All collected data, including subject responses, were automatically stored in Google Drive through Google Forms and Sheets.

2.3.2. Exhaustion Level

Exhaustion level refers to physical, mental, or combined fatigue [13]. While the Borg Rating of Perceived Exertion [14] and the Fatigue Severity Scale [15] assess physical exertion, they were deemed unsuitable for this study, as they focus on general physical work rather than wheelchair-related fatigue.

Instead, subjects rated their exhaustion levels as high, medium, or low through verbal feedback at the end of each session. The researcher recorded their responses by selecting the corresponding answer in a pre-set Google Form via a smartphone.

2.3.3. Workload

The subject’s workload was measured using the NASA Task Load Index (NASA-TLX), which defines workload as the effort required to achieve a certain performance level across six domains: mental demand, effort, frustration, performance, temporal demand, and physical demand [16]. Subjects rated these domains on a scale from 0 to 100, and their verbal responses were recorded in a pre-set Google Form via the researcher’s smartphone.

An online NASA-TLX tool (2019) was used to calculate the total workload score (0–100) and determine the contribution of each domain.

2.4.1. Questionnaire Administration

The data collection frequency was adjusted from once a week to once a month [10], while QoL data continued to be collected weekly. The researcher read the questionnaire aloud to the subjects and recorded their verbal responses in a pre-set Google Form on a smartphone, with the data automatically stored in Google Drive.

3.3.1. Wheelchair Usage

The data on wheelchair usage is presented in Table 2. Subject A completed 32 sessions and Subject B completed 29, with each session corresponding to one day. In total, both subjects completed 61 sessions. Subject A spent 1,722 minutes traveling 5,207 meters, while Subject B spent 1,492 minutes traveling 2,687 meters.

Subject A travelled 5,207 meters over 32 days (32 sessions), while Subject B covered 2,687 meters over 29 days (29 sessions).

3.3.2. Workload

The subjects' workload, assessed using the NASA Task Load Index, is summarized in Table 3. The six workload domains measured were mental demand, effort, frustration, performance, temporal demand, and physical demand.

For Subject A, the mean scores were:

• Lowest: Frustration (0.08), Physical demand (0.49), Temporal demand (0.50)
• Moderate: Mental demand (2.72)
• Highest: Effort (7.98), Performance (8.18)

For Subject B, the mean scores were:

• Lowest: Frustration (0.00), Temporal demand (0.63)
• Moderate: Mental demand (1.15), Effort (1.33)
• Highest: Performance (4.83)

Overall, the total workload for both subjects was low:

• Subject A: Mean score = 19.94
• Subject B: Mean score = 10.75

These results suggest that both subjects experienced a relatively low workload across the assessed domains.

4.2.1. Wheelchair Usage

The wheelchair usage was measured by the number of sessions, duration per session, and distance travelled.

• Subject A: 32 sessions, 1,722 minutes, 5,207 meters
• Subject B: 29 sessions, 1,492 minutes, 2,687 meters

Variations in distance and session durations were attributed to the subjects' declining health. However, both subjects remained enthusiastic, describing the wheelchair as “fun, new, and interesting.” The design allowed them to navigate freely without strict targets, aligning with studies that measure assistive device usage in terms of sessions, duration, and distance [24, 25].

4.2.2. Exhaustion Level

Exhaustion was rated on a three-point scale (low, medium, or high) based on verbal feedback from the subjects after each session. This approach mirrored a similar rating system used in a brain-controlled painting application study for ALS patients [10].

• Subject A: Reported 4 high-exhaustion sessions, with the remaining sessions rated as low or medium.
• Subject B: Reported 9 medium-exhaustion sessions, with the rest rated as low.

The ease of navigation using the eye-tracker device contributed to the overall low exhaustion levels.

Across 61 sessions, both subjects mostly reported low exhaustion (See Fig. 2):

• Subject A: Initially found the first 4 sessions highly exhausting but progressed to a lower exhaustion level over time.
• Subject B: Rated the first 9 sessions as medium exhaustion but later reported low exhaustion.

4.2.3. Workload

The NASA-TLX defines workload as “a hypothetical construct representing the cost incurred by a human operator to achieve a particular level of performance” [16, p.140]. Its main domains include mental, physical, and temporal demands, along with performance, effort, and frustration levels. In this study, NASA-TLX measured subjective workload after using the eye-controlled wheelchair.

Subject A reported a mean workload score of 19.94, while Subject B scored 10.75. For Subject A, performance (mean = 8.18) and effort (mean = 7.98) were the primary sources of workload, whereas Subject B experienced a higher physical demand (mean = 4.83) and performance pressure (mean = 1.33). Both participants identified as perfectionists, frequently adjusting the wheelchair until it reached their desired position. This suggests that their workload was influenced by personal goals rather than system limitations.

The low workload scores indicate that the wheelchair's user-friendly design contributed significantly to ease of use. Both participants noted that the wheelchair was easy to control and navigate with the aid of the eye-tracker. The system’s “bubble” feature, which provides on-screen visual feedback, enabled them to select directions by focusing and blinking. Notably, Subject A, a spectacle wearer, reported smooth performance, demonstrating the device's effectiveness for users with visual aids. The low workload and minimal exhaustion motivated both participants to engage regularly with the wheelchair.

Previous studies also support the effectiveness of assistive wheelchair technologies:

• A study [26] found that participants experienced a low workload while using an Intelligent Wheelchair System.
• A study [27] reported low NASA-TLX scores during trials with virtual reality-powered wheelchairs.
• Recent research [28] observed that users with tetraplegia experienced low-to-medium workload using a myoelectric control system.
• However, another study [29] noted that a mobility-enhancement robotic wheelchair induced a higher mental workload due to its complexity.

4.6.1. Reduced Physical and Emotional Burden

Compared to a non-motorized wheelchair, the ECW enhances patient independence by eliminating the need for upper limb strength to propel the wheelchair. This reduces the physical strain on caregivers who would otherwise assist in pushing a manual wheelchair. Additionally, by minimizing social exclusion, the ECW alleviates caregivers’ emotional burden, as they no longer feel responsible for the patient’s mobility limitations [40].

With the increased autonomy provided by the ECW, patients can move around freely, allowing caregivers more time to focus on other responsibilities rather than physical interventions, such as repositioning the wheelchair.

4.6.2. Improved Patient-Caregiver Interaction

The ECW allows caregivers and users to engage beyond basic mobility, facilitating more complex activities. Instead of focusing on manual assistance, caregivers can strengthen emotional bonds with patients by involving them in ADLs and instrumental activities of daily living. These activities may include grocery shopping, walking in the park, or caring for pets, thereby enhancing the patient's sense of self-worth while reducing caregiver frustration.

Seeing a patient regain independence and take control of his or her own life can be rewarding for caregivers, increasing their job satisfaction.

4.6.3. Challenges for Caregivers

Despite its benefits, the ECW presents new challenges for caregivers. They may still be responsible for operating, maintaining, and troubleshooting the wheelchair system. Older caregivers who are less familiar with computerized interfaces may experience anxiety over software errors or misconfigurations, which could endanger the patient.

Supervision is still required for eye-tracking errors, which could lead to unintended or unsafe movements. In cases of reduced eye endurance or system malfunction, caregivers may need to manually push the wheelchair. To enhance safety, a remote emergency button is available for caregivers, allowing them to assess the situation from a third-person perspective [40]. However, constant vigilance can increase caregivers' alertness and emotional stress.

Additionally, hired caregivers may face job insecurity as automated ECWs reduce the need for direct physical assistance.

4.6.4. Challenges Identified in ECW Research

A study on ECW users with amyotrophic lateral sclerosis (ALS) found that patients were generally satisfied with the independence provided by the device [41]. However, the study highlighted several challenges, including:

• Sunlight interference affects ocular tracking sensors, particularly in outdoor environments or under specific lighting conditions, such as direct sunlight or incandescent bulbs.
• Difficulty with reversing, with one-third of participants requesting a backup camera.
• Challenges for patients with strabismus who struggled to maneuver the ECW.
• Caregiver assistance is still required for vehicle transfers, as patients need help moving from the ECW into a car.

4.6.5. Striving for a Balance between Autonomy and Ethical Concerns

The increasing use of advanced technologies such as ECWs in healthcare raises ethical concerns, particularly regarding privacy, accessibility, and reliance on artificial intelligence for critical mobility functions.

4.6.6. Privacy Concerns

Eye-tracking data is classified as sensitive information, as it contains detailed insights into human attention, emotions, cognitive functions, preferences, and mental states [42]. This poses a significant risk to user privacy, especially if malicious service providers compile large databases to classify personal user data, including identity and gender, and exploit unauthorized data for marketing or surveillance purposes [42].

Another ethical concern involves the objectives of data protection systems. While these systems are designed to protect user data from third parties, privacy breaches occur if the system is used to influence and control users' thoughts and behaviours without their knowledge. Additionally, users may not always be fully aware of how developers, healthcare providers, or businesses collect, store, or use their data. Some systems may collect more data than necessary or use it for unintended purposes.

The balance between user privacy and system utility is entirely controlled by system providers. If this balance is not maintained, utility and profit may be prioritized over privacy, leading to potential exploitation. To mitigate risks, legal requirements should regulate the collection, storage, and use of highly sensitive eye-tracking data. Furthermore, strong cybersecurity measures must be enforced to safeguard user safety [43].

4.6.7. Accessibility and Socio-Economic Barriers

ECWs are expensive and are often inaccessible to individuals from a lower socio-economic group and developing countries. This creates a barrier to equal care and treatment for people with disabilities, particularly those in rural areas who have limited access to assistive technology and necessary medical support.

This lack of accessibility violates the ethical principle of distributive justice, which ensures equal access to healthcare resources for all individuals, regardless of their socio-economic status [44]. To promote fairness, ethical bodies should focus on reducing costs or offering subsidies for these technologies, potentially integrating them into public health programs.

4.6.8. Over-reliance on AI and Loss of Autonomy

Delegating critical healthcare tasks to AI systems may cause users to become overly dependent on technology, potentially reducing their autonomy and favouring machines over human caregivers. This can be particularly dangerous for individuals with cognitive impairments, as they may unknowingly issue hazardous commands to AI-controlled devices, putting themselves at risk.

Another ethical challenge is the loss of interpersonal interactions in healthcare. Replacing human caregivers with technology may lead to reduced trust and understanding, which is a widely debated ethical issue [45, 46].

4.6.9. Safety and Accountability in AI-Driven Systems

Safety is a major ethical concern when AI-powered systems interact directly with the physical world, such as autonomous vehicles, mobility devices like wheelchairs, or systems controlling essential healthcare services. If safety is compromised, determining who is responsible for failures becomes unclear [45]. Establishing clear accountability and ensuring rigorous safety testing are crucial to addressing this issue.