Visual fixation in disorders of consciousness: Development of predictive models to support differential diagnosis

Davide Sattin, Davide Rossi Sebastiano, Francesca Giulia Magnani, Ludovico D'Incerti, Giorgio Marotta, Riccardo Benti, Simone Tirelli, Anna Bersano, Dunja Duran, Elisa Visani, Stefania Ferraro, Ludovico Minati, Anna Nigri, Cristina Rosazza, Stefania Bianchi Marzoli, Paola Ciasca, Antonella Carcagni, Maria Grazia Bruzzone, Silvana Franceschetti, Matilde LeonardiDavide Guido

Research output: Contribution to journalArticlepeer-review


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.

Original languageEnglish
Article number113310
JournalPhysiology and Behavior
Publication statusPublished - Mar 1 2021


  • Diagnosis
  • Minimally conscious state
  • Predictive modeling
  • Vegetative state
  • Visual behavior
  • Visual system

ASJC Scopus subject areas

  • Experimental and Cognitive Psychology
  • Behavioral Neuroscience


Dive into the research topics of 'Visual fixation in disorders of consciousness: Development of predictive models to support differential diagnosis'. Together they form a unique fingerprint.

Cite this