1. Introduction
A tremor is a type of involuntary muscle contraction that results in shaking or trembling. It is a common symptom of movement disorders that affects a large number of people [
1]. There are various types of tremors, but the most prevalent ones are essential tremors (ET) and Parkinson’s tremors (PT). ET affects 0.9% of the population, whereas PT is estimated to affect 0.3% of individuals under the age of 60, increasing to 1% in those over 60 years old [
2]. ET is characterized by rhythmic body movements that have a frequency of 4–7 Hz [
3]. On the other hand, PT typically affects non-active body parts and tends to worsen when the limbs are straightened and placed in a resting position [
4].
Tremors can significantly impact an individual’s quality of life by affecting their ability to perform everyday activities such as eating, writing, and using a phone. These challenges can lead to feelings of frustration and a loss of independence. Therefore, it is crucial to invest in research and technological support to develop effective treatments, Rehabilitation, and support technologies that can alleviate the effects of tremors.
Pharmacotherapy has long been the conventional approach for managing tremors in patients with ET and PT, but its effectiveness is often hindered by various factors, including drug-drug interactions, intolerable side effects, and inadequate therapeutic response. For example, dopaminergic drugs, which are commonly prescribed for PT, only alleviate tremors in about half of the patients [
5]. Similarly, anticonvulsants such as gabapentin and paracetamol, as well as beta-blockers and benzodiazepines, which are often used for ET, have limited success rates, particularly for moderate to severe cases [
6,
7,
8]. Surgical interventions, such as deep brain stimulation, thalamotomy with radiofrequency, radiosurgery, and focused ultrasound, have shown limited long-term efficacy, which makes them a less popular option among patients [
9]. Furthermore, many individuals with mild tremors view medications as detrimental to their health, which leads them to abandon treatment, while those with mild to moderate tremors are generally hesitant to undergo surgery. For patients with severe and advanced tremors, surgery is often not a viable option due to the risks and complications associated with the procedure. Therefore, non-pharmacological interventions, such as regular exercise, must be considered as an alternative treatment approach.
Rehabilitation is an important adjunct to clinical medical treatment for tremors, particularly in patients with mild to moderate disease who are reluctant to undergo surgery and can benefit from physical therapy programs. Various approaches have been studied, including flexibility and strength training. Sajjad Farashi et al. [
10] conducted a meta-analysis of a large body of literature and found that exercise significantly reduced tremor in patients with Parkinson’s disease, with hand exercises showing promise for reducing distal limb tremors. Strength training involves applying resistance to the limbs of patients with tremors to stimulate neural control of muscles, and studies by G. Sequeira et al. [
11], J. Kavanagh et al. [
12], and M. Bilodeau et al. [
13] have shown that a generalized upper limb strength program has the potential to improve stability and flexibility in patients with PT or ET. However, these studies also found that patients may experience fatigue during training and that functional capacity does not always improve after training [
13]. Flexibility training is another useful method for improving execution and control in patients with tremors and gradually enhancing their self-confidence. Mona Kadkhodaie et al. used eccentric-based rehabilitation training and found a significant reduction in the amplitude of resting tremor after exercise in the intervention group, although the study had limitations in terms of assessing tremor fluctuation [
14]. N.E. Vance et al. [
15] used yoga to rehabilitate patients with primary tremor and demonstrated improvement in tremor assessment scales after an eight-week intervention. W. Chung et al. [
16] provided behavioral relaxation training to ET and PT patients and found that regular relaxation training can reduce the effects of tremor, but noted the limitations of general rehabilitation training due to a lack of long-term follow-up. H. Rajalin [
17] found that home rehabilitation training can improve not only the physical function of patients with Parkinson’s disease but also their activities of daily living, but noted the lack of interventions available for use at home by patients with PT. Overall, rehabilitation has the potential to improve the quality of life of patients with tremors, but further research is needed to develop effective and accessible interventions for all types of patients. In order to promote daily rehabilitation for Parkinson’s patients and reduce the pressure on rehabilitation institutions, Xiangya Hospital in China has developed a video exercise program called Yapa-PBGA (Yapa Parkinson balance and gait aerobics). This exercise program increases muscle control, improves gait and balance disorders, and upper limb flexibility in Parkinson’s patients [
18]. It is cost-effective and convenient for patients to perform daily rehabilitation at home. However, its limitations are that this video format cannot provide direct rehabilitation training guidance and monitoring to patients, especially considering the uncontrollable situation of Parkinson’s patients. It is difficult to achieve the ideal training effect through this training method.
Virtual reality or augmented reality technology can overcome the limitations of video-based rehabilitation training methods by providing interactive visuals and personalized guidance, thus increasing patient engagement and motivation during the rehabilitation process. Hueso et al. [
19] developed a virtual reality remote assistance system that allows therapists to create customized treatment plans and automatically record the patient’s movement. However, the system lacks flexibility and is not suitable for developing tremor rehabilitation training. J. Cornacchioli [
20] studied the use of the Oculus Rift grip as a tool to detect Parkinson’s symptoms by measuring involuntary hand movements. The study tested the effectiveness of Parkinson’s symptoms and concluded that the accuracy of Oculus Rift was sufficient for measurement needs but did not develop the technology for tremor rehabilitation training. G.C. Burdea et al. [
21] reviewed the advantages and disadvantages of virtual reality rehabilitation applications and summarized the benefits, including improved patient motivation, adaptive data access, online data access, and reduced healthcare costs. They also revealed the limitations of virtual reality technology in rehabilitation, including a lack of supportive infrastructure, expensive equipment, and inadequate communication infrastructure for rural tele-assistance. Jiang et al. [
22]. presented a multi-category gesture recognition model that uses signals from both surface electromyography and inertial measurement units. The model aims to improve the accuracy and robustness of gesture recognition in various real-world applications, such as human-computer interaction and rehabilitation.
Augmented reality refers to the combination of digital information from a virtual world with the physical environment of the real world to create a more interactive and immersive user experience. Wang et al. [
23,
24] conducted a study on the application of augmented reality technology to assist tremor patients in typing on a keyboard like ordinary people. Wang et al. [
25] investigated the use of projection-based augmented reality technology called “Extend Hand” to help tremor patients directly interact with remote-controlled home appliances. Compared to virtual videos, augmented reality can provide more intuitive and specific rehabilitation training guidance and monitoring. By adding digital elements to real-life scenes, patients can better understand, simulate, and practice rehabilitation training skills, thus enhancing their rehabilitation outcomes. In addition, augmented reality technology can provide more personalized rehabilitation training plans and feedback, which can enhance patient engagement and motivation, leading to more active participation in rehabilitation training. Therefore, augmented reality technology has great potential to become an effective auxiliary tool for rehabilitation training. Aditya Pillai et al. [
26] developed an innovative mixed reality rehabilitation tool specifically designed for upper limb injuries that utilized HoloLens 2 technology. The tool overlaid digital elements onto the real-world scene through the augmented reality feature of HoloLens 2, providing personalized rehabilitation training guidance and monitoring to help patients with upper limb injuries regain their function. However, the expensive price of HoloLens 2 limits its adoption for rehabilitation training that can be conducted at home by users. The Oculus Quest 2 by Meta is a milestone virtual reality headset that allows for free movement without the constraints of cables and is available at an affordable price. Its passthrough feature can be developed into an AR device. Additionally, the built-in controller Oculus Touch, featuring infrared emitters and inertial sensors, has been proven to be accurate and robust in detecting hand movements [
27,
28].
Considering the lack of a specialized augmented reality system for rehabilitation training specifically designed for patients with tremors and the need to improve the effectiveness and experience of rehabilitation training that can be conducted at home by patients with tremors, including one-on-one demonstration, posture guidance, and monitoring of training progress, we developed a low-cost augmented reality tremor rehabilitation training system using Oculus Quest 2. This study used Optitrack to record standard Yapa-PBGA movements and incorporated them into the system, creating a one-on-one virtual model for rehabilitation training. By designing natural interaction logic in the system, individuals with tremors can be guided to perform accurate limb movements during rehabilitation training. The patients’ actual limb movement data can also be recorded and analyzed in real-time by the system.
5. Results
As a way of clearly describing, analyzing, and comparing the two rehabilitation modalities in the following sections, “Video” and “AR” represent the video and AR rehabilitation conditions, respectively, “Q1” to “Q10”, the experience investigations, and “G1” to “G9” (“L” is the left body, “R” is the right body), the rehab postures.
A compared T-test was conducted to compare the magnitude of the mean body movement of participants making G1 to G9 under the “video” and “AR” conditions (see
Figure 6). There was a significant effect on the mean body movement magnitude at the
p < 0.05 level for G2, G3, G4, G5, G7, and G9. In groups G1 and G8, significant differences were commonly found in the L group, but not found in the R group. The results revealed that for most rehab postures, the magnitude of human motion in the AR condition was significantly greater than that in the video environment. As shown in
Figure 6, the red lines indicate the actual magnitude of movement of the control group in making each rehab posture, respectively. From the results of the experiment, the AR group was remarkably closer to the control group compared to the video group. To sum up, the proposed AR method can help individuals with tremors make Yapa-PBGA postures that achieve a magnitude of body movements close to that of a standard demonstrator and can effectively promote the quality of movements in rehabilitation training for them in comparison to the video.
Some psychological aspects of the survey further evaluated participants’ experience with “Video” and “AR” rehabilitation training. A compared T-test was used to analyze the difference in emotional experience between the conditions of “Video” and “AR” through the questions from Q1 to Q5. The results are given in
Figure 7. Both conditions showed statistically significant differences with
p < 0.01. The results showed that the participants felt comfortable and easy doing rehabilitation, did not perceive the physical burden, and were able to tolerate the intensity of the rehabilitation training. In the analysis, we compared the differences between the two conditions on the basis of Q6 to Q8 and found significant differences at
p < 0.01. The result shows that the participants were more likely to be guided by rehabilitation training. Thus, the participants were more confident in performing standard rehabilitation training movements in the AR environment. There are no significant differences between the conditions in Q7, however, as the participants generally rated it very well (See
Figure 7), it means participants could follow the pace of rehabilitation training under the AR condition. The analysis of Q9 and Q10 utilized a comparative T-test, revealing that the participants generally expressed satisfaction with the AR environment and preferred it as a rehabilitation training setting.
In addition, the proposed methods were evaluated from the perspective of a sense of ownership and agency. The sense of ownership (SO) and the sense of agency (SA) are two central aspects of bodily self-awareness; the sense of ownership would be considered a direct perceptual experience, while the sense of agency is the sense of controlling and causing the body to act through volition, two important properties of operational logic in human-computer interaction. As we calculated the mean and standard deviation on the questionnaire of Q11 (SO) and Q12 (SA), we found a mean of 5.5 with a variation of 0.48, and a mean of 5.7, with a deviation of 0.3. This implies that participants are prone to the illusion that they are touching virtual objects, that the interaction can be easily controlled, and that users do not experience a strong sense of dissonance in the AR environment.
6. Discussion
Rehabilitation is an important way to reduce the severity and progression of tremors, and Yapa-PBGA provides a low-cost video rehabilitation program that can be used by patients at home. However, this program has limitations in providing face-to-face guidance for patients and does not track the quality of movements during rehabilitation. To address these limitations and better assist tremor patients with their rehabilitation at home, a low-cost AR system has been demonstrated to create an immersive rehab experience. This system simulates an instructor demonstrating gymnastics in front of the user, while also guiding the user’s body movements during the rehabilitation process.
To evaluate the effectiveness of the proposed AR system in supporting individuals with tremors to achieve better results in Yapa-PBGA gymnastics, experiments were conducted to compare and analyze the magnitude of body movements of individuals with tremors in the video and AR environments, as well as that of the control group. The magnitude of movement is an objective measure of how well the movement achieves the correct posture, and the greater the magnitude, the better the rehabilitation will be. In most rehabilitation postures, participants demonstrated significantly greater magnitudes of limb movement in the AR environment, approaching the demonstrator’s level, compared to the video environment. However, no statistically significant differences were observed in G1 (R) and G8 (R). Further analysis revealed that this was due to the fact that these two groups used poses involving hands on the waist, which are easy to assume and least influenced by experimental conditions. Therefore, G1 (R) and G8 (R) were not considered in the final experimental conclusion. It can be concluded that the proposed AR method is effective in improving the quality of movement for individuals with tremors. User experience surveys showed that participants felt more relaxed, comfortable, and engaged when using the proposed AR system compared to the video environment. The proposed AR system was also effective in providing rehabilitation guidance to individuals with tremors. Although there was no significant difference between the AR environment and the video environment in Q7, the actual evaluation revealed that participants could follow the pace of rehabilitation training under AR conditions. Additionally, an analysis of the sense of ownership and agency of participants with tremors found that the AR system was effective in providing them with a natural rehabilitation experience without causing dissatisfaction.