In Comatose Patients, MR Shows Radical Restructuring of Brain Networks

An international group of scientists have studied data from 17 comatose patients using functional magnetic resonance imaging (fMRI) information. Their research revealed that the brain networks of these patients have been restructured. These findings could help clinical practitioners to optimize the diagnosis of comatose patients.

The researchers are focused on studying brain networks of brain-damaged comatose (nontraumatized) patients, a state where the individual is considered to be unconscious. The findings were published December 2012 in the journal Proceedings of the National Academy of Sciences of the United States of America (PNAS). The authors of the study employed a novel graph theory-based methodology, where images were assembled using functional MRI data at rest and using statistical signal-processing technology. Local and overall effectiveness indices of functional brain networks were gathered for 17 brain-injured patients and 20 healthy study participants. Correlations in 417 brain areas were obtained to create brain connection graphs using the statistically significant correlations.

Researchers from Inserm (Institut National de la Santé et de la Recherche Médicale) unit 836 Grenoble Institut des Neurosciences (France), CNRS (Centre National de la Recherche Scientifique; Paris, France), the GIPSA lab (Grenoble, France), and the Behavioural and Clinical Neuroscience Institute (Cambridge, UK), working with clinical practitioners from the Strasbourg University Hospital Centre (France), have been able to highlight restructured brain networks in brain-damaged (nontraumatized) comatose patients.

Through comparisons with the healthy subjects, the findings revealed that the overall cerebral connectivity is preserved in comatose patients. By examining the connectivity at a local level, the authors of the study have observed that some brain regions (“hubs”), which are very connected in healthy volunteers, are less well connected in comatose patients. On the other hand, the less densely connected regions in the healthy subjects’ networks become “hubs” in comatose patients.

The connectivity graph method is used to summarize in a single image data acquired through MRI scanning. It converts the effectiveness of connections in a single region compared to all the others. By grouping the most interconnected regions, modules are revealed (each represented by a different color). Patients and healthy volunteers both have different models in their spatial location, representing radical changes to the brain connections.

According to current hypotheses, consciousness disorders in persistently comatose patients could be linked to disconnection phenomena between specific cortical regions, in particular the precuneus. The study’s findings also point in this direction. “From an overall perspective, the topology of brain connections resists well to traumatism by reorganizing the most interconnected regions in the network. It therefore seems that comas may be linked to changes in the location of ‘hubs’ among the brain networks,” suggested Chantal Delon Martin, an Inserm researcher.

Patients with brain injury may go through a range clinically-defined states: a vegetative state that is characterized by the preserved sleep-wake cycle (i.e., eyes opening spontaneously, autonomous breathing); minimally conscious state where patients have partially preserved environmental consciousness (eye movement capacity, reaction to stimulation); locked-in syndrome where the patient is paralyzed but conscious (communication using eyes); brain death when the coma is irreversible flat line EEG [electroencephalography], no blood flow).

Brain injuries evaluations in comatose patients are currently performed through clinical examination, morphologic MRI, evoked potentials and by SPECT (single-photon emission computed tomography (SPECT) positron emission tomography (PET). “The results of this study could help clinical practitioners in the difficult diagnosis process for comatose patients, since this method makes it possible to characterize each patient individually,” concluded the researchers.

Related Links:
Grenoble Institut des Neurosciences
Inserm GIPSA Lab
Behavioral and Clinical Neuroscience Institute