Clinical validatiOn of an AI-based DEcision Support System for Robotic Upper Limb Rehabilitation in Patients With Stroke. A CO-AIDER Study. (COAIDER)

Don Carlo Gnocchi Onlus Foundation

Status: Not yet enrolling

Conditions: Stroke; Stroke Hemorrhagic; Stroke Ischemic; Upper Limb Rehabilitation

Treatments: Device: Robotic upper limb rehabilitation in patients with Stroke without AI support; Device: Robotic upper limb rehabilitation in patients with Stroke with AI support

Study type: Interventional

Funder types: Other

Identifiers

NCT07199322

COAIDER

RF-2021-12375472 (Other Grant/Funding Number)

Details and patient eligibility

About

The goal of this clinical trial is to validate an artificial intelligence-based decision support system (AI-DSS) for robotic rehabilitation in participants with stroke.

The study aims to answer the following questions:

• Can the AI-DSS support therapists in setting therapy parameters and adapting personalized robotic rehabilitation programs effectively?
• Can AI-supported robotic rehabilitation improve upper limb function, activities of daily living (ADL), cognitive function, and quality of life compared with standard therapist-guided robotic rehabilitation?

Researchers will compare two groups:

• Robotic rehabilitation supported by the AI-DSS (CO-AIDER);
• Robotic rehabilitation with parameters set by therapists (control group).

Participants will:

• Receive robotic rehabilitation for the upper limb, either guided by the AI-DSS or by therapists;
• Be monitored throughout the program, with therapy parameters adjusted according to their progress;
• Complete assessments to evaluate changes in upper limb function, activities of daily living, cognitive function, and quality of life.

The study will also evaluate the usability and acceptability of the AI-supported robotic system as perceived by participants and clinical staff, and will include a cost-effectiveness analysis of the two interventions.

Full description

The CO-AIDER study is a multicenter, randomized, controlled, two-arm, non-profit clinical trial validating an AI-based decision support system (AI-DSS) in robotic upper limb rehabilitation for stroke patients. It supports therapists in setting therapy parameters and personalizing rehabilitation programs while monitoring patient progress. A total of 100 participants with ischemic or hemorrhagic stroke in the subacute phase (within six months of the event) will be randomly assigned to two groups. The experimental group will receive upper limb robotic rehabilitation using the MOTORE device supported by the AI-DSS CO-AIDER, which predicts motor recovery and provides recommendations for optimal therapy parameters based on clinical and instrumental data. Recommendations focus on stiffness, weight, and viscosity, key parameters for tailoring therapy to patients' abilities. The system provides guidance only and does not directly affect therapy progression. The control group will receive the same robotic treatment, with parameters set by therapists based on their clinical expertise. All participants will complete 30 sessions of robotic rehabilitation, 45 minutes per day, five days per week, alongside a conventional rehabilitation program for lower limbs and trunk. Assessments will be performed at baseline (T0, within 7 days of enrollment), mid-treatment (T1, after 15 sessions), and at the end of treatment (T2, after 30 sessions). Clinical assessment will include measures of upper limb motor function, activities of daily living, pain, spasticity, cognitive function, and quality of life using standardized clinical scales. Robotic performance will be assessed using parameters collected by the MOTORE system. Usability and acceptability of the AI-DSS will be evaluated through validated questionnaires. Randomization will be stratified by disease onset, motor impairment, and recruitment center to ensure balanced allocation. Data will be collected and stored pseudonymously in a secure, encrypted database. A cost-effectiveness analysis will compare AI-DSS-supported rehabilitation with therapist-guided therapy. The study will also evaluate how patients and clinical staff perceive the AI-supported robotic system in routine rehabilitation.

Enrollment: 100 estimated patients

Sex: All

Ages: 18 to 85 years old

Volunteers: No Healthy Volunteers

Inclusion criteria

• Diagnosis of ischemic or hemorrhagic stroke confirmed by Computed Tomography (CT) or Magnetic Resonance Imaging (MRI);
• Age between 18 and 85 years;
• Time since stroke less than six months;
• Mild to severe upper limb impairment (Fugl-Meyer Assessment for Upper Extremity (FMA-UE) ≤ 58).

Exclusion criteria

• Behavioral or cognitive disorders and/or reduced compliance;
• Severe spasticity or hypertonia in the affected limb (Modified Ashworth Scale (MAS) > 3);
• Severe visual impairments;
• Concurrent participation in another clinical trial for upper limb rehabilitation following stroke;
• Refusal to provide written informed consent.

Trial design

Primary purpose: Other

Allocation: Randomized

Interventional model: Parallel Assignment

Masking: Double Blind

---

100 participants in 2 patient groups

Robotic treatment with an AI-based decision support system (Robotic therapy + CO-AIDER software)

Experimental group

Description:

Participants assigned to this arm will receive upper limb robotic treatment using the MOTORE robotic device, supported by the AI-based decision support system (CO-AIDER software). CO-AIDER implements a predictive model based on recurrent neural networks (LSTM) that analyzes demographic, clinical, and kinematic data collected during treatment sessions. The model estimates the expected post-treatment scores on three standard clinical scales (FMA, ARAT, and MI). In addition to outcome prediction, the system provides therapists with recommendations on robot parameters (stiffness, weight, and viscosity), which are used to personalize the difficulty level of the exercises implemented in MOTORE, without directly altering the therapeutic course.

Treatment:

Device: Robotic upper limb rehabilitation in patients with Stroke with AI support

Robotic treatment with parameters defined exclusively by the clinical team (Robotic Therapy)

Active Comparator group

Description:

Participants assigned to the control group will receive upper limb robotic treatment using the MOTORE robotic device. The treatment will be identical to the experimental arm in terms of device and exercise types, but all parameters will be determined exclusively by the physiotherapists based on their clinical expertise, without AI-based decision support.

Treatment:

Device: Robotic upper limb rehabilitation in patients with Stroke without AI support