From Preference to Performance: Patient-Centered Design of Multimodal Cueing in Parkinson’s Disease Gait Training

Paper Title

From Preference to Performance: Patient-Centered Design of Multimodal Cueing in Parkinson’s Disease Gait Training

Publication Info

• Topic area: Patient-centered design for Parkinson’s disease rehabilitation using multimodal cueing.
• Keywords: Parkinson’s disease, gait training, multimodal cueing, wearable technology, rehabilitation, patient-centered design, visual cueing, auditory cueing, somatosensory cueing, CARES framework.

Background and Problem

• Problem / challenge: Current rehabilitation for Parkinson’s disease (PD) gait disorders heavily depends on clinicians and clinic-based interventions, which are costly, inaccessible, and burdensome. Existing systems often neglect early-stage PD needs, gait stability, and patient preferences.
• Significance: Addressing these gaps could provide continuous, home-based training, reduce healthcare costs, and improve patient independence and quality of life.
• Motivation and related work: Prior studies show that visual, auditory, and somatosensory cueing can improve gait disorders, but they lack systematic exploration of modality-specific effects and user preferences. Existing systems are clinician-driven, with limited patient involvement in design.

Solution

• Proposed approach: A lightweight, patient-centered rehabilitation system that uses inertial measurement units (IMUs) to sense gait and generate multimodal cueing (visual, auditory, somatosensory) for training.
• Novelty:
  1. Development of the CARES framework: Centered on the patient, Adaptive and personalized, Reinforced and guided, Evidence-based, and Safe.
  2. Introduction of two cueing paradigms (periodic and process) and six multimodal cueing conditions.
  3. Comparative evaluation of cueing types and modalities in user studies with early-stage PD patients.
  4. Integration of patient feedback into system design for personalized and adaptive rehabilitation.
• Procedure and key techniques:
  1. Conducted interviews with 10 early-stage PD patients and 5 clinicians to identify needs and challenges.
  2. Designed a system incorporating IMUs, augmented reality, open-ear headphones, and vibration modules.
  3. Evaluated the system in two user studies with 16 PD patients, comparing cueing types (periodic vs. process) and modalities (visual, auditory, somatosensory).

Results

• Concrete findings:
  • Process cueing improved gait velocity and stability more than periodic cueing for visual and auditory modalities (e.g., 14.75% vs. 2.43% improvement in auditory velocity).
  • Periodic cueing was more effective for somatosensory stimulation (e.g., 3.47% vs. 1.21% improvement in velocity).
  • Auditory cueing produced the largest objective improvements (e.g., 18.21% velocity increase), while visual cueing had the highest subjective preference and lowest workload.
• Advantage over baselines:
  • All cueing conditions significantly outperformed the no-cueing baseline in gait velocity (e.g., 9.15% improvement for process cueing) and stability (e.g., Gait Consistency Index increased from 57.15% to 68.67% with process visual cueing).
  • Patients preferred visual process cueing (mean preference score: 4.5/5) despite auditory cueing achieving better objective outcomes.
• Experiments / evaluation:
  • User Study 1: Compared periodic vs. process cueing across modalities with 6 patients.
  • User Study 2: Compared visual, auditory, and somatosensory cueing with 16 patients.
  • Metrics: Gait velocity, Gait Consistency Index (GCI), Relative Difference Index (RDI), NASA-TLX workload, and patient preferences.
• Limitations and future work:
  • Limited to early-stage PD patients with preserved cognitive and visual function.
  • Short experimental period in controlled environments.
  • Future work includes extending accessibility features, testing with advanced PD patients, and conducting long-term, in-home deployments.

Summary

This study developed a patient-centered, multimodal cueing system for Parkinson’s disease gait rehabilitation, guided by the CARES framework. Two user studies demonstrated that process cueing is more effective for visual and auditory modalities, while periodic cueing suits somatosensory input. Auditory cueing achieved the best objective outcomes, but visual cueing was preferred by patients due to lower workload and perceived companionship. These findings highlight the importance of balancing performance and user acceptance in rehabilitation design, paving the way for personalized, home-based training systems.