The loss of upper-limb motor functioning due to ischemic stroke has a negative impact on quality of life of patients. Action Observation Treatment (AOT), recently developed based on the mirror mechanism functioning, has proved effective in promoting plasticity into the motor system, ultimately boosting the motor recovery. The use of virtual reality further empowers the effectiveness of AOT, making the visual experience more first-person, and thus more realistic. A complete picture of the neurobiological mechanisms underlying AOT effectiveness is lacking to date, including the neuroradiological and biohumoral markers modulated by AOT, and their role in predicting the clinical outcome. Starting from these premises, the investigators propose a study aiming at evaluating the AOT effectiveness in post-stroke patients, and at identifying the biomarkers indexing the motor recovery process.

Sixty patients affected by ischemic stroke will be enrolled at Humanitas Clinical and Research Center IRCCS (ICH) Stroke Unit will be included and randomized in the following groups:

• VR-AOT: experimental group, observing actions in virtual reality
• VR-LO: control groups, observing a matched dose of videos depicting landscapes in virtual reality.

Clinical, neuroimaging and biohumoral evaluations, performed at Screening (Ts), Basal visit (T0), end-of-treatment (T1), 2 months follow up (FU-2), will include the following endpoints: Fugl-Meyer upper extremity motor scale (FM-UE); Nine-hole-peg test; Box and block test; Modified Ashworth scale; MRC; Pain Numeric-rating-scale; Functional Independence measure.

Patients will undergo treatment sessions for 6 weeks (5 days/week). VR-AOT patients will observe in virtual reality -rehearse and execute specific upper limb motor task belonging to activities of daily living. VR-LO participants will observe virtually explorable landscapes lasting for a matched duration. Then, participants will be asked to actively perform the same set of hand actions requested to VR-AOT experimental groups.

An intention-to-treat (ITT) analysis will be performed, using rmANOVA, minimal clinically important difference (MCID) in clinical outcomes, and chi-square test. For the primary endpoints, a two-arms rmANOVA with time as within-subject factor (3 levels: T0, T1, FU2) and group (VR-AOT vs VR-LO) as between-subjects factor will be conducted. The same approach will be adopted for secondary outcome measures. Subsequently, the effect of clinical, neuroradiological and biohumoral baseline features will be tested as regressors on the primary outcome variations via linear or ranked regression models. Significance will be set at 5%, and adjusted for MC.

Inclusion Criteria:

• Age>18
• Ischemic stroke in the territory of middle cerebral artery (MCA, lesion extension lower than half of its territory) or lacunar stroke
• MRC score 2-4 in at least one segment of the affected limb
• Modified Rankin Scale (mRS) prior to stroke ≤ 2
• Able to perform study requirements
• Able to give informed consent according to ICH/ GCP, and national/local regulations

Exclusion Criteria:

• Presence of global aphasia
• History of seizures
• Posterior circulation stroke
• Significant ipovisus
• Moderate-to severe neglect
• Cognitive impairment or language barriers
• Psychiatric comorbidities
• Drug or alcohol abuse
• Autoimmune disease
• Contraindication to perform MRI