THE PREDICTOR PROJECT

The PREDICTOR Project:

Explainable Predictive Models for Stroke Rehabilitation

The PREDICTOR Project — Explainable Predictive Models for Stroke Upper Limb Robot-Based Rehabilitation — aims to develop and validate an innovative technological framework that integrates robotics and wearable sensors for monitoring recovery in stroke patients with upper limb motor disabilities.

Using Machine Learning and Explainable AI, the project seeks to:

Improve post-stroke rehabilitation management;
Support clinicians’ decisions through interpretable, data-driven models;
Personalizing therapy to each patient’s specific condition and stage of recovery, by providing optimal adjustments to the robot–patient interaction parameters.

To predict motor recovery, a combination of data from the following will be used:

Clinical assessment scales;
The MOTORE robotic system;
Inertial Measurement Units (IMU);
Wearable sensors, to collect data in both clinical and home environments.

This integrated approach enables a more accurate evaluation of each patient’s progress, support to therapeutic decisions and an improvement of the effectiveness of rehabilitation, anticipating its outcomes.

Understanding the Challenge: Stroke and Motor Recovery

Picture displaying the arm of a doctor managing data on a tablet. The picture is taken from above, and a desk is on the background of the photo, with a medical device and some notes placed on it

Stroke is one of the leading causes of death and disability worldwide, with a significant social and healthcare impact.
According to the World Health Organization (WHO), stroke is the second leading cause of death and the third leading cause of disability globally.

In Italy, approximately 200,000 new cases of stroke occur each year, and about 80% of patients experience motor impairments affecting the upper limbs. Despite the use of intensive rehabilitation programs, only a small percentage of patients fully regain function in the affected arm or hand.

This highlights an urgent need for innovative approaches that leverage new technologies to improve motor recovery and tailor rehabilitation programs to each individual’s needs.

Innovation in Robotic and Digital Rehabilitation

Picture displaying the arm of a patient using the robotic rehabilitation device "Motore". In the picture you can also see the screen where the game he is playing with the device is shown

In recent years, research in robotic rehabilitation and digital health technologies has opened new perspectives for the treatment of post-stroke motor deficits.
Moreover, Artificial Intelligence (AI) and Machine Learning (ML) are emerging as powerful tools for analyzing complex data and creating predictive models that can support clinical decision-making.

Robotic systems allow therapists to deliver rehabilitation that is more intensive, repetitive, and personalized compared to traditional physiotherapy. Devices such as the MOTORE system can monitor biomechanical parameters and provide quantitative insights into patient progress.

Meanwhile, wearable sensors make it possible to monitor patients at home, collecting continuous data on motor and physiological conditions.
In addition, devices such as inertial measurement units (IMUs) and biometric wrist sensors track movement, physical activity, and vital signs, offering a full picture of the patient’s recovery, both in clinical and daily-life contexts.

Finally, Explainable Artificial Intelligence (XAI) introduces a new paradigm: models that are transparent and interpretable, helping clinicians understand why an AI makes a specific prediction and supporting trustworthy, data-driven care.

Funding

PREDICTOR has been selected and funded under the PRIN 2022 program (Progetti di Ricerca di Rilevante Interesse Nazionale) by the Italian Ministry of University and Research (MUR).
This recognition underscores the project’s high scientific and technological value and its relevance to advancing research in robotic and AI-assisted neurorehabilitation.

The project brings together several leading Italian institutions:

Politecnico di Bari
University of Milano-Bicocca
University of Bari Aldo Moro
University of Foggia

By combining expertise in bioengineering, robotics, digital medicine, and data science, the PREDICTOR partners aims to develop cutting-edge solutions that will improve the quality of life for stroke survivors.

PARTNERS

Page updated

Report abuse