The visual fixation represents a doubtful behavioral sign to discriminate Vegetative from Minimally Conscious State (MCS). To disentangle its meaning, we fitted univariate and multivariable logistic regression models matching different neurophysiological and neuroimaging data of 54 patients with Disorders of Consciousness to select the best model predicting which visual performance (visual blink or pursuit) was shown by patients and the best predictors set. The best models found highlighted the importance of the structural MRI and the visual evoked potentials data in predicting visual pursuit. Then, a qualitative pilot test was made on four patients showing visual fixation revealing that the obtained models correctly predict whether the patients' visual performance could support/correlate to a cognitively mediated behavior. The present pilot models could help clinicians to evaluate if the visual fixation response can support the MCS diagnosis.

Distinguishing Vegetative from Minimally Conscious State (VS; MCS) is a challenge due to the difficulty in relating observed behaviors with underlying brain activity. The visual responses in the visual subscale of the Coma Recovery Scale-revised (CRS-r) are informative of the change from VS to MCS; however, few studies explored the functional and structural integrity of the visual system in these patients.

One of the major challenges for clinicians who treat patients with Disorders of Consciousness (DoCs) concerns the detection of signs of consciousness that distinguish patients in Vegetative State from those in Minimally Conscious State. Recent studies showed how visual responses to tailored stimuli are one of the first evidence revealing that one patient is changing from one state to another. This study aimed to explore the integrity of the neural structures being part of the visual system in patients with DoCs manifesting a reflexive behavior (visual blink) and in those manifesting a cognitively and cortically mediated behavior (visual pursuit). We collected instrumental data using specialized equipment (EEG following the rules of the International 10-20 system, 3T Magnetic Resonance, and Positron Emission Tomography) in 54 DoC patients. Our results indicated that visual pursuit group showed a better fVEPs response than the visual blink group, because of a greater area under the N2/P2 component of fVEPs (AUC could be seen as an indicator of the residual activity of visual areas). Considering neuroimaging data, the main structural differences between groups were found in the retrochiasmatic areas, specifically in the right optic radiation and visual cortex (V1), areas statistically less impaired in patients able to perform a visual pursuit. FDG-PET analysis confirmed difference between groups at the level of the right calcarine cortex and neighboring right lingual gyrus. In conclusion, although there are methodological and theoretical limitations that should be considered, our study suggests a new perspective to consider for a future diagnostic protocol.