Abstract
Worldwide, between 20-57% of strokes result in vision loss or partial cortically-induced blindness. The consequences of this condition are not only perceptual but also visuomotor; it is known that saccades in these patients are biased towards their blind field. Since this condition normally impairs extrafoveal vision, it’s often assumed that fixation is unaffected and that fixational eye movements are normal. To test these assumptions, we measured fixational eye movements in 11 patients (age=56±4yrs, 2F/9M) and 10 aged-matched control observers (age=45±4yrs, 6F/2M) using a high-precision digital Dual-Purkinje Image eyetracker. Under monocular viewing, in alternative trials, participants either performed an active foveal 4AFC high-acuity discrimination task, or a passive fixation task in which they maintained the gaze on a central 16’x16’ target for 5s. Stimuli in the acuity task consisted of isolated digits presented for 500 ms, and their size was adjusted using an adaptive procedure. Our results show that fixation precision was comparable between patients (Bivariate-Contour Ellipse Area [BCEA] =3.2±0.20deg²) and controls (BCEA = 2.9±0.17deg²). However, while controls’ gaze was centered on the fixation target during the passive task, patients exhibited significant offsets in gaze position. Offsets were present throughout the fixation duration and were systematically biased toward the blind field, with an average shift of 4.2’±1.1’ (unpaired t-test, p = 0.024), but they disappeared when patients performed the high-acuity foveal task. Patients and controls also had similar visual-acuity thresholds, indicating that gaze offsets recorded in the passive fixation task were not due to reduced foveal vision. These findings show that cortical blindness affecting extrafoveal vision introduces micro-offsets in gaze position during passive fixation. We posit that this abnormal behavior may reflect an attentional bias toward the blind field when patients are not engaged in a foveal task.