The goal of this clinical trial is to assess the usability and feasibility of a robotic and virtual reality-based system called NeglectARm 3D for the three-dimensional assessment of unilateral spatial neglect (USN) in patients with brain injury due to ischemic or hemorrhagic stroke.

The main questions it aims to answer are:

• Is the NeglectARm 3D system usable and feasible for patients with USN?
• Do the robotic measurements show preliminary reliability and validity compared with conventional clinical neglect assessments (e.g., BIT, CBS)?

Participants will:

• Perform two assessment sessions, three weeks apart (T0 and T1);
• Interact with the NeglectARm 3D system using a robotic end-effector and a VR headset to reach sequentially presented targets in a 3D virtual space;
• Receive robotic assistance if targets are not reached, allowing the system to record trajectories, spatial positions, and assistance applied;
• Complete standard clinical neglect tests to enable comparison between robotic and conventional measures.

This study aims to develop a repeatable, standardized, and objective tool for the assessment of USN in patients with brain injury, potentially improving evaluation of depth perception and spatial attention deficits.

This observational, non-commercial study evaluates the usability and feasibility of the NeglectARm 3D system, a robotic and virtual reality-based tool for three-dimensional assessment of Unilateral Spatial Neglect (USN) in patients with brain injury due to ischemic or hemorrhagic stroke. Participants undergo two assessment sessions (T0 and T1), spaced three weeks apart, which include clinical neglect assessments using standard instruments and robotic evaluations with two test-retest trials per session.

The robotic component is the Franka Research 3 arm, a force-sensitive manipulator with seven degrees of freedom and torque sensors in each joint, providing industrial-quality pose repeatability of ±0.1 mm and minimal path deviation at high speeds. The virtual reality component is the Meta Quest 3 headset, which delivers high-resolution, smooth, and immersive visual feedback and tracks hand movement in real time using integrated sensors and RGB cameras.

During each session, participants reach sequentially presented stimuli in a 3D virtual space using their unaffected hand, in contact with the robotic end-effector. The system measures three-dimensional kinematics and provides assistance if the participant cannot reach a target independently. The robot operates in shared control mode, remaining transparent when the participant can reach a target autonomously ("patient-guided") and guiding the participant along the optimal trajectory when assistance is required ("robot-guided"), ensuring safe and natural movement.

The system records spatial positions and assistance parameters, enabling the construction of a map of target interactions and extraction of parameters for instrumental assessment of USN. Between assessment sessions, participants continue conventional rehabilitation. Data management includes checks for consistency and range, with source data verification performed by comparing robotic measurements with recorded observations. The study plans to enroll 15 patients, each completing two assessment sessions, to evaluate usability, feasibility, and preliminary reliability and validity of the system.

Inclusion Criteria:

• Age > 18 years;
• Diagnosis of ischemic or hemorrhagic stroke;
• Diagnosis of Unilateral Spatial Neglect (assessed by The Hearts Test, Oxford Cognitive Screen, cut-off ± 2);
• Signed informed consent form.

Exclusion Criteria:

• Presence of language deficits (inability to complete the OCS test);
• Presence of cognitive decline prior to the stroke event;
• Presence of severe visual and/or hearing impairments;
• Language barriers;
• Orthopedic, neurological, or psychiatric conditions that could interfere with the study or with the assessment of motor and cognitive function;
• Uncontrolled epilepsy despite antiepileptic treatment;
• Behavioral disorders that may interfere with study participation;
• Presence of medical devices whose functioning could be compromised (e.g., pacemaker, etc.);
• Object-centered neglect (assessed by the OCS Hearts Test; object asymmetry ± 2).

The enrollment phase will be conducted at the IRCCS Fondazione Don Carlo Gnocchi in Florence (FI) and at the Fondazione Don Carlo Gnocchi in Sant'Angelo dei Lombardi (AV), in Italy. Each patient will be considered enrolled after assessment of inclusion and exclusion criteria and after signing the informed consent form. A total of 15 patients will be recruited, with 10 participants at the Florence center and 5 participants at the Sant'Angelo dei Lombardi center.