Learning a Shape-adaptive Assist-as-needed Rehabilitation Policy from Therapist-informed Input

2025-10-06

AI Summary

To address insufficient safety and poor adaptability in current therapist-robot collaborative rehabilitation, this paper proposes a teleoperated, on-demand assistive rehabilitation framework. The method encodes real-time corrective forces applied by therapists as via-points in a latent space and employs a shape-adaptive artificial neural network (ANN) policy to enable partial, progressive, and dynamic deformation of reference trajectories—thereby jointly respecting patient movement preferences and therapeutic intent. It integrates latent-space modeling, ANN-driven rehabilitation control, teleoperation, and online trajectory adaptation. Evaluated on two representative rehabilitation tasks, the framework significantly reduces corrective force magnitude (average reduction: 32.7%) and improves motion smoothness (jerk reduced by 28.4%) compared to state-of-the-art approaches, thereby enhancing personalization, safety, and remote applicability.

Technology Category

• Intelligent Robots: Manipulation
• Natural Language Processing: Safety and Robustness
• Planning, Routing, and Scheduling: Replanning and Plan Repair

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Search and Retrieval-Augmented AI: Assisted, interactive, and conversational search

Abstract

Therapist-in-the-loop robotic rehabilitation has shown great promise in enhancing rehabilitation outcomes by integrating the strengths of therapists and robotic systems. However, its broader adoption remains limited due to insufficient safe interaction and limited adaptation capability. This article proposes a novel telerobotics-mediated framework that enables therapists to intuitively and safely deliver assist-as-needed~(AAN) therapy based on two primary contributions. First, our framework encodes the therapist-informed corrective force into via-points in a latent space, allowing the therapist to provide only minimal assistance while encouraging patient maintaining own motion preferences. Second, a shape-adaptive ANN rehabilitation policy is learned to partially and progressively deform the reference trajectory for movement therapy based on encoded patient motion preferences and therapist-informed via-points. The effectiveness of the proposed shape-adaptive AAN strategy was validated on a telerobotic rehabilitation system using two representative tasks. The results demonstrate its practicality for remote AAN therapy and its superiority over two state-of-the-art methods in reducing corrective force and improving movement smoothness.

Problem

• Developing shape-adaptive assist-as-needed policy for rehabilitation
• Encoding therapist input into latent space via-points
• Deforming reference trajectory based on patient preferences

Innovation

• Encodes therapist force into latent space via-points
• Learns shape-adaptive assist-as-needed rehabilitation policy
• Deforms reference trajectory using patient motion preferences