Personalized Assistance in Robotic Rehabilitation: Real-Time Adaptation via Energy-Based Performance Monitoring
Bibliographic record
Abstract
Recent studies underscore the importance of the patient’s active contribution and voluntary effort in enhancing therapy outcomes in physical rehabilitation. This paper presents an adaptive control scheme to implement active robotic rehabilitation. The primary goal is to dynamically regulate robotic assistance based on the patient’s performance and individual conditions, encouraging active participation, and effective therapy. To achieve this, a Lyapunov-based adaptive algorithm is developed that dynamically adjusts the admittance parameters by balancing the error and effort minimization. A novel performance index based on human energy input enables real-time identification of the intended human sharing role. This index is used as an adaptive rate in the proposed algorithm to enhance the control system’s dynamic responsiveness to changes in human performance. The proposed approach achieves two main rehabilitation objectives. First, it encourages active and safe human participation. Second, it enhances the therapy by providing personalized assistance, tailored to individual abilities and conditions, and thus reduces the need for therapist intervention. The performance of the proposed approach is illustrated in experimental studies. The results demonstrate the adaptability of the algorithm, ensuring compliant and safe interaction and effective task completion. Note to Practitioners—In a human-robot cooperation (HRC) framework, the automatic adaptation of the robot’s role as well as safe and stable interaction are crucial. These aspects are amplified in the context of robotic rehabilitation due to the special conditions of the human participants. Classic control methods, in shared control, lack system intelligence and automation in role allocation. However, the shared role of humans in HRC, particularly in rehabilitation applications, introduces real-time and unpredictable variations. This study addresses the shortcomings of classic control methods, by integrating intelligence into the control system through an adaptive Neural Network algorithm in shared autonomy. To emulate human-like adaptability, two crucial aspects are considered. Firstly, it incorporates safety assurance embedded in the adaptive algorithm via Lyapunov-based adaptation. Secondly, it detects the human’s role within the control loop through a novel energy-based performance index, which views the human as an active contributor to the system’s dynamic energy flow. This ensures robust behavior by dynamically adjusting the trade-off between task completion and minimal robot intervention. A standout feature of our algorithm lies in its expendability to exoskeleton systems, making it highly versatile for use in robotic rehabilitation and assistive technologies. The algorithm’s design allows for straightforward integration with exoskeletons, requiring only interaction force measurements in the joint space. It facilitates monitoring of a patient’s performance in each joint using the proposed performance index based on the human energy entry into the system. Beyond rehabilitation, the algorithm’s ability to adjust autonomy levels through adaptation makes it applicable to a wide range of Human-Robot Cooperation scenarios where automatic role allocation is necessary. Preliminary experiments underscore the adaptive algorithm’s robust responsiveness to changes in human performance. Future investigations should involve clinical experiments addressing real-life challenges associated with various movement deficiencies and responding to real-time issues that may arise during rehabilitation sessions.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".