- The paper demonstrates that adding wrist abduction-adduction assistance significantly improves ADL performance, with marked increases in spill-free drinking and effective scratch stick leveling.
- The methodology integrates a tongue-based interface with a low-inertia, tendon-driven wrist actuation module, enabling precise and user-specific control within a 6-DoF exoskeleton.
- The study highlights that targeted functional wrist ROM is more critical than maximal range, enhancing usability and task execution in assistive exoskeletons.
Clinical Assessment of Wrist Abduction-Adduction Assistance in 6-DoF Upper-Limb Exoskeletons for ALS and SCI
Introduction
This paper presents a clinical evaluation of the functional and user-perceived impact of an additional wrist abduction-adduction (Ab-Ad) joint in a tongue-controlled, six-degree-of-freedom (6-DoF) upper-limb exoskeleton (EXOTIC2) for individuals with amyotrophic lateral sclerosis (ALS) and spinal cord injury (SCI) (2604.20967). While most exoskeleton designs prioritize proximal joints and often ignore comprehensive wrist motion—specifically radial-ulnar deviation—this study explicitly investigates the clinical relevance and usability of including wrist Ab-Ad support during performance of activities of daily living (ADLs).
System Architecture and Experimental Methods
EXOTIC2 builds on its predecessor with an additional tendon-driven wrist Ab-Ad DoF, complementing shoulder, elbow, forearm, and grasp assistance. Grasp is achieved through a cable-driven soft robotic glove, and joint parameters are individually configured for user-specific passive or active ROM. The wrist actuation module employs a low-inertia, proximally-mounted design to minimize bulk at the distal segment.
Tongue-Based Interface
Control of the full exoskeleton, including the new wrist DoF, is achieved via a non-invasive intraoral tongue-computer interface (nTCI) with 15 inductive sensors. Command selection relies on spatial activation of tongue gestures mapped to a GUI, supporting 3 one-dimensional sliders plus a two-dimensional joystick for proportional actuation velocity, with safety overrides and visual feedback.
Cohort and Protocol
Six participants (1 ALS, 5 SCI) with severe upper-limb deficits were enrolled, all previously trained on the nTCI. Each subject performed standardized drinking and scratch stick leveling ADLs, with and without wrist Ab-Ad assistance, in a randomized within-subject design. Objective metrics included task completion time (TCT), success rate, task-specific kinematics, and a structured usability questionnaire.
Results
Enabling wrist Ab-Ad led to robust, consistent improvements in ADL execution:
- Drinking Task: Spill-free completion rate increased from 16.7% to 72.2% with wrist Ab-Ad enabled, while overall spillage events decreased (from 77.8% to 22.2%). Task failures remained unchanged at 5.6%.
- Scratch Stick Leveling: Successful leveled placements improved from 16.7% to 77.8%. Failed placements fell from 66.7% to 16.7%.
- Task Completion Time: Four of six participants exhibited TCT reductions of 10–75 seconds for the drinking task when using the additional wrist DoF. Scratch stick TCTs showed minor reductions or were stable, reflecting lower complexity.
Kinematic Analysis
Analysis of event-based wrist angles revealed task- and strategy-dependent usage. For drinking, grasp was associated with higher adduction, shifting towards abduction at release. In scratch stick tasks, most releases converged on positive (abducted) wrist angles. Across all users, only a subset of their total available ROM was engaged during task execution. This affirms that the ability to target functional regions of motion is more critical than maximizing ROM.
User Feedback
Questionnaire data demonstrated:
- Low discomfort ratings for both interface and exoskeleton use (all medians 1–4 on a 10-point scale, low scores are favorable).
- Strongly perceived functional benefit from the wrist Ab-Ad joint, especially in the drinking task.
- No increase in discomfort with additional DoF.
- Varied acceptance related to donning/doffing effort and ergonomics, with two participants noting lower willingness to use the system, primarily due to the control interface rather than the wrist DoF itself.
- No user reported worsened performance with the wrist enabled, and overall acceptance improved for half the participants.
Discussion
The clinical data substantiate that integration of wrist Ab-Ad actuation in ULEs meaningfully enhances task success in ADLs requiring precise end-effector orientation and object manipulation. Objective gains were mirrored in user perception, particularly for tasks with alignment constraints (drinking). Importantly, the inclusion of the new DoF did not exacerbate physical discomfort or interface burden, supporting its integration from a human-robot interaction perspective.
The finding that successful task execution relies on precise and controllable access to specific regions of motion, rather than on maximal ROM, aligns with prior rehabilitation robotics literature. It further implies that exoskeleton design optimization should focus on functional workspace and intuitive control, rather than mechanical range.
Variability in usability responses underscores the need for further work on donning/doffing, overall ergonomic refinement, and tailored control schemes—as the acceptance and adoption in real-world settings hinges on the seamless integration of engineering capability with user-centered priorities.
Implications and Future Directions
This research advances the clinical evidence base supporting comprehensive wrist actuation, especially Ab-Ad, in ULEs for users with severe paralysis. The results indicate that technical additions must be systematically validated for their real-world, user-centered utility rather than assumed to be beneficial due to increased DoF.
Future development should prioritize:
- Personalization of wrist ROM based on individual task needs.
- Enhancements in control interfaces for increased intuitiveness and reduced cognitive load.
- Ergonomics to optimize device donning/doffing, fit, and comfort in daily environments.
- Larger and more diverse participant cohorts, including substantial ALS representation, for quantitative generalization.
Conclusion
The study demonstrates that wrist abduction-adduction assistance in a tongue-controlled 6-DoF exoskeleton significantly improves functional performance in representative ADLs for individuals with ALS and SCI. The evidence advocates for the inclusion of selective distal DoFs in assistive exoskeletons, contingent on rigorous user-centered validation, with a strong emphasis on usability and task relevance as determinants of clinical utility (2604.20967).
