Abstract
The study of disorders of consciousness (DoC) is very complex because patients are suffering from a wide variety of lesions, affected brain mechanisms, different symptom severity and are unable to communicate. Combining neuroimaging data and mathematical modeling can help us quantify and better describe some of these alterations. This study’s goal is to provide a novel analysis and modeling pipeline for fMRI data leading to new diagnosis and prognosis biomarkers at the individual patient level. To do so, we project patient’s fMRI data into a low dimension latent-space. We define the latent space’s dimension as the smallest dimension able to maintain the complexity, non-linearities, and information carried by the data, according to different criteria that we detail in the first part. This dimensionality reduction procedure then allows us to build biologically inspired latent whole-brain models that can be calibrated at the single-patient level. In particular, we propose a new model inspired by the astrocyte regulation of neuronal activity in the brain. This modeling procedure leads to two types of model-based biomarkers (MBBs) that provide novel insight at different levels: (1) the connectivity matrices bring us information about the severity of the patient’s diagnosis, and, (2) the local node parameters, correlate to the patient’s etiology, age and prognosis.
Graphical abstract: The BOLD fMRI time series (magenta) are projected into a latent space of low dimension (purple) using an auto-encoder (left). In latent space, different latent whole-brain models are fitted to each subject’s data (center). The fitted model parameters constitute 1st-level MBBs, allowing the differentiation between diagnosis (blue, magenta, and cyan). Finally, transverse clusters based on the fitted models can be related to the patient’s etiology, age, and prognosis (right).
Competing Interest Statement
The authors have declared no competing interest.
