During the active phase, patients will receive real-time audio feedback on spontaneous alpha-band functional connectivity between ipsilesional associative visual areas and the rest of the brain. This will allow them to learn to improve their pathological brain interactions. The neurofeedback training will last about 40 minutes, with frequent breaks. It will be followed by visual stimulation of the affected visual field according to recommendations for inducing steady-state visual evoked potentials.

The control phase is structured identically to the active period, except that the acoustic neurofeedback is synthetically generated and not linked to the subject's actual functional connectivity, while still resembling its dynamic characteristics to ensure effective blinding. The training will last about 40 minutes, with frequent breaks. The training will be followed by visual stimulation of the affected visual field, just as in the active condition.

The proposed neurofeedback approach relies on high-density electroencephalography (EEG) combined with advanced source localization algorithms. Data will be analyzed in real-time and simultaneously recorded for offline analysis. During each update, a data segment will be filtered between 1 and 20 Hz. The beamformer, computed at the beginning of the session, will be used to project the signal to the gray-matter voxels.

The investigators will compute the alpha-band absolute imaginary coherence between a visual target area and the rest of the brain as index of functional connectivity. Global functional connectivity in the alpha band (8-13 Hz) between the voxels in the target region and the rest of the brain will be calculated.

This will be evaluated using the Haag-Streit Octopus 900 perimetry device (Haag-Streit AG, Köniz, Switzerland). The device features advanced gaze-tracking capabilities that effectively control compensatory eye movements, ensuring accurate measurement of visual field improvements. The Central 30-2 protocol will be followed.

The primary outcome will be the change in the Mean Deviation (MD) score in detection threshold (in dB) from baseline to the end of the intervention period within the target area of affected visual field. The MD score represents the overall deviation of the patient's visual field from age-matched normative data, with more negative values indicating greater visual field loss.

Neurophysiological changes reflected in alpha-band network communication between targeted area and the rest of the brain, as measured with EEG-based FC. For this, the investigators will record resting state EEG, which, together with individual MRI scans and beamforming techniques, will allow us to estimate FC in source space.

Reaction time in the TAP Visual Scanning test is a continuous variable with no predefined bounds, with longer reaction times indicating poorer performance; omissions represent the number of missed targets, with higher values also indicating poorer performance.