Multimodal assessment of coma and development of fMRI-based methods to study consciousness in the rat

Abstract

Despite the continuous research on human coma, some critical questions remain unanswered. For instance, why does the brain become unconscious? what is the reason why some patients awake from a coma while others remain in a persistent unresponsive syndrome? how does the brain recover from this radical unconscious state? The limitations inherent to human research (e.g. multiple etiologies, unfeasible timing, etc) harden identification of the key elements underlying the pathophysiology of coma and its recovery. The first goal of this PhD thesis is to provide a new platform whereby studying coma in a systematic and reliable manner is possible. We produced a model of brainstem coma in rats by causing brainstem ischemia. Once the procedure was established, I assessed the acute restoration of brain function using whole brain functional Magnetic Resonance Imaging during rest (rs-fMRI). Multi-modal analysis converged into the identification of a network composed of basal forebrain, basal ganglia and thalamus that seemed to participate in the reactivation of the cerebral cortex during early neurological recovery after coma. Methodologically, imaging the non-paralyzed rodent (e.g. non-anesthetized comatose animals) presents the challenge of identifying and eliminating potential motion artifacts derived from spontaneous breathing. To clarify this issue, in a parallel line of study, I imaged the brain of non-paralyzed anesthetized animals under different ventilatory rates and MRI repetition times. Spectral analysis of the acquired signals revealed a pronounced interference of the respiration on the fMRI data, which can be predicted and controlled, to an extent, by choosing appropriate ventilatory settings and sampling rates. Another important limitation of studying coma through fMRI in small animals is the fact that no behavioral assessment can take place inside the MRI scanner. To overcome this constraint, I established a multimodal fMRI platform that includes pupillometry to allow tracking arousal fluctuations in parallel to the MRI scanning. Besides setting up the platform for future animal fMRI studies assessing consciousness, this work identified a series of regions which oscillatory function co-vary with pupil dilations, which may constitute the first pupil-driven arousal network identified in rats and provides a way of tracking pupil-governed brain state changes from the fMRI data. Last, in view of a potential work merging fMRI with stimulation of deep brain structures like the hypothalamus, which appeared overshadowed in the rat coma study, I contributed to implementing an MRI-compatible robotic arm to precisely position optical fibers (for stimulation or recording) into the rat brain. This tool has proven to increase the successful rate and reliability of deep brain targeting in fMRI studies. This thesis aims at providing a better understanding of the loss and recovery of consciousness by working with the rat as an animal model and fMRI as the main imaging modality. In addition, it attempts to resolve some of the challenges that emerge while performing brain imaging in small animals and presents novel platforms for the investigation of neural mechanisms related to arousal.