Neurological symptoms such as muscle weakness, loss of motor, sensory and cognitive are common in stroke patients. These symptoms result in limitations in functional daily activities such as gait ability and stair climbing [1]. Among these, abnormal gait is characterized by increased energy expenditure and asymmetry, contrasting with the smooth and coordinated limb movements of normal gait. Abnormal gait often presents as reduced gait speed, shortened stride length, and an asymmetric weight bearing [2]. This asymmetry increases the risk of secondary injuries, such as falls, and imposes significant restrictions on daily living[3]. Consequently, gait recovery is one of the primary goals of rehabilitation therapy for stroke patients. Recent studies have explored various approaches to promote functional gait recovery.

Among these approaches, task-specific training provides an environment that enables patients to focus on targeted movements and tasks with motivation [4]. Robot-assisted gait training (RAGT) is a representative example [5]. RAGT facilitates effective gait training by guiding the lower limb movements of patients according to pre-programmed, normal physiological gait patterns [6]. This programmed robot-assisted gait training uses normal gait patterns to promote proper joint movements, timing of the gait cycle, and adequate weight support, ultimately enhancing gait symmetry [7]. Research has shown that such symmetric gait training can induce neuro-plastic changes through repetitive learning [8]. Additionally, studies have demonstrated that combining traditional physical therapy with RAGT improves gait speed and enhances gait independence in stroke patients [9]. Research on robotic rehabilitation has become sophisticated and specialized. Recent studies reveal that as motor function improves in stroke patients, changes occur in cortical activation, suggesting that motor recovery is related to the reorganization of motor neural networks. Cortical activation is enhanced when various task-specific rehabilitation exercises are implemented [10,11]. Based on this evidence, sensory input plays a critical role in improving balance ability and gait ability in stroke patients, leading to the development of new intervention methods utilizing sensory stimulation. The central nervous system, composed of complex and dynamic parallel connections, demonstrates hierarchical and rhythmically synchronized movements. However, a stroke disrupts this rhythmicity, resulting in asymmetric gait patterns. Consequently, rhythmic auditory stimulation training, which synchronizes and guides the brain’s motor and perceptual areas through rhythmic external stimuli, has gained prominence as an intervention [12,13]. Previous studies using RAS have shown not only improvements in stride length and gait speed but also enhancements in gait symmetry [12,14]. Furthermore, studies by Ford et al. [15] and Roerdink et al. [16] found that treadmill training with RAS improved upper and low-extremity coordination and increased pelvic and thoracic rotation. These findings underscore the positive effects of RAS on walking ability. Given this evidence, RAS and robotic- assisted gait training are hypothesized to complement each other effectively. Therefore, this study aims to investigate the effects of RAS-based robotic-assisted gait training on the balance and gait of stroke patients.

Participants

This study included a pre-posttest control group design where the subjects were divided according into intervention methods, such as the Auditory stimulation with Robot-assisted Gait Training group (ARGT) or Over-Ground Gait training (OGG) group. The subjects of the study were selected as 40 patients who were admitted to Chung-Nam national university hospital in Daejeon. To minimize the selection bias, the following selection criteria were applied randomly to the two group. The inclusion criteria were as follows: (1) diagnosis of stroke (minimum 6 months post-stroke) (2) ability to walk 10 meter (without assistive device) (3) balance ability (maximum Berg Balance Scale score 45) (4) cognitive abilities enabling communication (minimum Mini Mental State Examination score 24) The general characteristics of the participants are shown in figure 1.

Intervention

The participants of this study were 40 patients selected from those admitted to Chung Nam national university hospital in Daejeon for physical therapy, based on specific inclusion criteria. Participants in the study were divided into two groups using a first-come, first-served lottery system and trained. Before initiating the training, participants’ balance and gait abilities were assessed, and they were divided into the ARGT and the OGG. The robot-assisted gait training group received gait therapy using the Lokomat Pro (Hocoma AG, Zurich, Switzerland), an exo-skeletal type robotic device. The ARGT underwent robot-assisted gait training three times per week, 30 minutes per session, for 6 weeks. The OGG group also trained for the same period and frequency. General physical therapy consisted of a 30 minute per session for six weeks.

ARGT (Auditory stimulation Robot-assisted Gait Training)

The ARGT utilized rhythmic auditory cues in the form of regular metronome beats to guide gait training. The intervention was tailored to each patient’s comfortable gait speed by adjusting the metronome tempo to match their individualized steps per minute. The robot-assisted gait training involved measuring each patient’s steps per minute and using this data to conduct gait training at their stable gait speed. During the 3rd and 5th weeks of the training program, the metronome tempo was increased by 5% according to each patient’s capacity, promoting gradual improvement. To acclimate participants to the robot-assisted gait training program, they were allowed a one-minute practice session to familiarize themselves with the metronome tempo. Following this, they engaged in 30minute gait training sessions synchronized with the metronome beats. The training protocol lasted for 6 weeks, with sessions conducted three times per week, for 30 minutes per session.

Outcome Measures

1. Medical Research Council (MRC)

The MRC was applied to evaluate lower extremity muscle strength. The MRC is divided into six grades: Normal (5), Good (4), Fair (3), Poor (2), Trace (1), Zero (0). Hip flexion, extension, abduction, knee flexion, extension, ankle dorsiflexion, plantar flexion of affected side lower extremity was assessed and has a total score of 30 points. The mean value was recorded.

2. Berg Balance Scale (BBS)

In this study, the BBS was used as a balance assessment tool. The BBS are 14 different items that evaluates the degree of the balance and fall risk in stroke patients. The evaluation items are for dynamic and static balance. And it takes about 15minutes. The evaluators were therapists with more than three years of clinical experience and conducted assessment before and after intervention.

3. Timed Up and Go test (TUG)

The TUG is a representative test that can evaluate dynamic balance. Therefore, in this study, TUG was conducted and a total of three attempts were made and the average value was recorded.

4. 10 Meter Walking Test (10MWT)

The 10MWT was used to evaluate walking ability and walking speed. The 10MWT was conducted and a total of three attempts were made and the average value was recorded.

5. Fugl-Meyer Assessment (FMA)

In this study, the FMA, a representative motor function assessment, was evaluated. FMA evaluated lower extremity motor function, excluding upper extremity function, and has a total score of 34 points.

6. Modified Barthel Index (MBI)

The MBI evaluates whether patients with brain damage can perform activities of daily living, and serves as the basis for functional judgment of sequelae. Therefore, in this study, the MBI was conducted to determine the impact on daily life through functional recovery.

Data Analysis

The statistical analysis of this study was performed using IBM SPSS Statistics for Windows, version 25.0 (IBM Corp., Armonk, NY, USA). The general characteristics of the subjects were tested for normality using descriptive statistics (Shapiro-Wilk test). Paired t-tests were used to evaluate differences before and after treatment. Also, Independent t-tests were performed to determine the significance of differences between groups. The significance level was set at p < 0.05.

General characteristics and medical characteristics of subjects

The general characteristics and medical characteristics of all subjects in the ARGT and OGG groups were all homogenous (Table 1).

Changes in balance and gait ability of the participants

Balance and gait abilities were assessed with BBS, TUG and 10MWT. Also muscle strength and functional restoration were assessed with MRC, FMA and MBI. The MRC, BBS, TUG, 10MWT, FMA and MBI significantly increased in the ARGT and and OGG groups post-intervention (p<0.05). The total MRC score (Post-Pre) was significantly improved in ARGT (3.41±0.90) compared to OGG (2.31±0.60). Similarly, total BBS score (Post-Pre) was significantly improved in ARGT (7.75 ±4.00) compared to OGG (5.12±1.50). Also total TUG score (Post-Pre) was significantly improved in ARGT (-10.08±5.17) compared to OGG (-6.62±3.44). The MRC, BBS and TUG of the outcome measures showed a greater significant increase in ARGT compared to OGG (p<0.05) (Table2). But the 10MWT, FMA and MBI of the outcome measure showed no significant improved in ARGT compared OGG (p>0.05) (Table3).

This study aimed to compare robot-assisted gait training with rhythmic auditory stimulation (RAS) and over-ground gait training to improve gait performance in relation to changes in balance ability in 28 stroke patients over 6 weeks. It also sought to propose an effective exercise method to enhance functional recovery in stroke patients.

Post stroke patients often exhibit asymmetric balance abilities and impaired postural control, leading to significant postural sway during weight shifting and an increased risk of falls, which raises the likelihood of secondary injuries [17,18]. To minimize such asymmetry in gait, this study employed the Lokomat Pro, an exo-skeletal type robot that strives to balance gait discrepancies between the affected and unaffected sides. Robotic devices are typically categorized into exo-skeletal type robot and end-effector type robot. The Lokomat Pro, a representative exo-skeletal type robot, mimics human joint movements and promotes ideal extremity motion during walking. The device integrates a treadmill with a body weight support harness, allowing patients to train in a fall-safe environment with focused task-oriented programs. Many prior studies on stroke patients applied auditory stimuli with fixed tempos [14,19,20]. However, recent studies on Parkinson’s and stroke patients demonstrated significant improvements in walking speed when auditory stimuli were applied at varied tempos (90%, 100%, 110% of stable walking speed). Faster tempos particularly resulted in more significant increases in walking speed [16,21-23]. Based on these findings, this study incrementally increased the tempo of auditory stimuli by 5% per week relative to the patient’s measured stable walking speed [13]. The experimental period was designed based on studies such as Thaut et al. [13], which reported significant gait improvements with a regimen of 3 weeks, 5 sessions per week, and 30 minutes per session. And Whitall et al. [20] reported upper limb strength improvements with a protocol of 6 weeks, 3 sessions per week, and 20 minutes per session. To further enhance training outcomes, this study implemented a protocol of 6 weeks, with 3 sessions per week, lasting 30 minutes per session.

Muscle weakness is one of the most prominent deficits caused by motor impairments in stroke patients. Such muscle weakness is a key factor limiting functional recovery in stroke patients, with ankle muscle weakness being particularly problematic as it affects both balance and gait abilities [24]. Wolfson et al. [25] found that loss of balance ability in stroke patients is closely related to ankle muscle weakness, often due to spasticity in the plantar flexors. This spasticity prevents the heel from making contact with the ground during the stance phase, leading patients to walk on the forefoot or toes. Consequently, the stance phase shortens, push-off becomes difficult, and during the swing phase, foot drop occurs, where the toes drag along the ground. These abnormalities result in slower walking speed and inefficient gait patterns. Ultimately, stroke patients exhibit asymmetric balance abilities, which increase their risk of falls [17,18]. To improve movement efficiency in stroke patients, strengthening the lower limbs is essential for restoring balance and gait abilities [25]. Thaut et al. [12] demonstrated that RAS reduces variability in the EMG patterns of the gastrocnemius muscle in stroke patients. In a subsequent study, Thaut et al. [26] reported that RAS not only reduced variability in muscle activation but also improved gait speed, symmetry, and smoothness. Excessive stretch reflexes and spasticity are major contributors to muscle weakness in stroke patients. Robot-assisted gait training has been recognized as an effective method for managing these issues. For instance, Mayr et al. [6] reported a significant reduction in lower limb spasticity after robot-assisted gait training. Based on this evidence, this study aimed to examine changes in lower extremity strength that could influence balance ability. Lower extremity strength was assessed using the MRC. The ARGT showed a significant increase in lower extremity strength pre-post treatment, with a mean change of 3.41±0.90 (p 0.05). Moreover, < when compared to the OGG, the ARGT demonstrated significantly greater improvements, confirming that robot-assisted gait training is a more effective method for enhancing muscle strength. These findings support the potential of robot-assisted gait training as an effective therapeutic approach for addressing muscle weakness in stroke rehabilitation.

Balance refers to the ability to maintain posture and stability while supporting body weight without falling. It is one of the most critical components of physical recovery for stroke patients [27]. Balance involves stabilizing the body against gravity and other external forces. The ability to maintain balance in a standing position is closely related to the walking ability of stroke patients [28]. Particularly, the balance ability of the affected lower limb in stroke patients is strongly correlated with their walking and functional ability [29]. Thus, improvements in balance and walking abilities are likely interrelated. This study evaluated changes in balance using the BBS for static balance and the TUG for dynamic balance. The BBS scores, representing static balance, showed significant improvements in both the ARGT and the OGG after the intervention (mean change: 7.75±4.00) (p<0.05). However, between-group comparisons revealed that the ARGT achieved significantly greater improvements than the OGG (p<0.05). Similarly, the TUG results, which reflect dynamic balance, demonstrated a significant reduction in completion time in the ARGT after the intervention (mean change: 10.08±5.17) (p <0.05). Between-group comparisons also showed that the ARGT outperformed the OGG (p<0.05). These findings suggest that robot-assisted gait training with RAS is more effective in improving both static and dynamic balance abilities in stroke patients compared to over-ground gait training. Ultimately, it can be inferred that robot-assisted gait training using auditory stimulation led to the transfer of balance ability through muscle strength recovery. Lower extremity strength and balance ability are closely related to recovery of gait [24].

Normal walking involves balanced and coordinated limb movements, requiring minimal energy for efficiency. Walking stability is regulated by automatic strategies integrating visual, proprioceptive, and vestibular systems [30]. However, sensory impairments caused by stroke can lead to spasticity and altered muscle responses, forcing patients to rely on residual muscle function for movement. This results in asymmetrical weight distribution, reduced weight-shifting ability, and shorter stride lengths during walking [31]. Such asymmetry destabilizes walking and reduces gait speed, ultimately leading to energy inefficiency and functional limitations [32]. These gait deficits can significantly affect patients’ daily activities, reducing their independence and causing social restrictions [33]. Thus, improving gait ability is a key goal in rehabilitation for stroke patients [31]. In this study, gait improvements were measured using the 10MWT, and functional recovery was assessed through the FMA and MBI. Post stroke recovery was related to two main mechanisms (True recovery, compensation). Among these, true recovery refers to brain reorganization closely related to neuroplasticity. This recovery process is associated with repeated motor learning. Repeated motor learning is critical for true recovery, encompassing skill acquisition, motor adaptation, and decision making [34]. Treadmill training, a widely used method in stroke rehabilitation, is based on motor learning theories and has shown to be effective in improving gait patterns through repetitive practice [35]. However, the physical demands on therapists and the potential for inconsistent assistance during training have led to increased interest in robot-assisted gait training [36]. Robot-assisted gait training with RAS combines rhythmic sensory cues with repetitive gait practice. RAS activates both auditory and motor pathways via subcortical auditory processing systems and cortical regions, enhancing movement synchronization and symmetry [37]. Studies by Dias et al. [38], Mayr et al. [6], and Jung et al. [39] have shown that robot-assisted gait training improves gait ability and endurance. Furthermore, Wong et al. [40] and Bonnyaud et al. [41] demonstrated that robotic gait training improves dynamic balance and gait symmetry. RAS has been shown to further enhance gait symmetry and temporal measures, such as stance phase duration and symmetry index [42]. The 10MWT revealed significant improvements in walking speed for the ARGT, with a mean pre-post intervention change time (10.33±4.83 seconds) (p 0.05). While < the OGG also showed improvements (mean change: 7.81±2.71 seconds) (p <0.05), between-group comparisons did not yield statistically significant differences (p>0.05). Similarly, functional recovery measured by FMA and MBI showed significant improvements in both groups pre-post intervention (p<0.05), but no significant differences between groups (p>0.05). These results suggest that while ARGT positively impacts walking ability, its superiority over OGG is less clear. In addition, these results suggest that robot-assisted gait training using auditory stimulation is less effective than over-ground gait training as a method for improving walking ability, and thus is insufficient to induce functional changes. The absence of significant between-group differences may be attributed to the short training duration, small sample size, and uncontrolled variables during the study.

Robot-assisted gait training with RAS demonstrated positive effects on lower limb strength and balance in stroke patients. However, its superiority over ground gait training in improving walking remains inconclusive. Future research should address these limitations by employing larger sample sizes, controlling for external variables, and conducting follow-up studies to evaluate long-term effects. Additionally, optimizing training protocols to maximize the synergistic effects of robotic assistance and rhythmic auditory stimulation is recommended.

The authors declared no potential conflicts of interest with respect to the authorship and/or publication of this article.