Disorders of consciousness (DoC) are a group of disorders that can occur after severe brain injury. DOC have been subdivided based on behavioural observations into: Coma, lacking any signs of wakefulness or awareness; the vegetative state, showing signs of wakefulness but lacking any signs of awareness; and the minimally conscious state, showing signs of wakefulness and infrequent and irregular signs of awareness. The so-called locked-in syndrome, a state where both wakefulness and awareness are intact, but no communication is possible due to a lack of muscle function, does not belong to the disorders of consciousness. However, it is difficult to distinguish the locked-in syndrome from DoC diagnostically, because consciousness can only be shown through consistent responses to a command and current methods for assessing consciousness rely on behavioural responses. Patients with locked-in syndrome might not be able to move voluntarily at all in the most severe cases. Behavioural assessment would then classify them as unaware. While this is an extreme and rare case, it illustrates the problem behavioural assessment poses. Such assessments are unable to distinguish the effects of impaired muscular control from impaired awareness, when either has reached an extreme level of severity. Brain damage that does not affect consciousness itself can nevertheless affect the results of the behavioural assessment of consciousness. It is then hardly surprising that the diagnosis of DoC is associated with a high level of uncertainty. The advantage of brain imaging methods is that they do not rely on the patients ability to produce a consistent behavioural response. There have therefore been efforts to use the brain imaging methods electroencephalography, positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) to aid diagnosis of disorder of consciousness. PET and fMRI have successfully been used to identify regions of difference in some patients in a DoC. Task-based fMRI has been used to identify intact consciousness, using tasks that require explicit understanding of instructions and wilful modulation of brain activity, but no motor control. One of these tasks consists of periods where the participant imagines playing tennis alternated with periods of rest. The ability to follow this paradigm is evidence of consciousness, and a few patients with a diagnosis of DoC have been shown to be able to do this task. However, the tennis task requires high order processing of the tasks requirements and the majority of patients does not respond to this task. fMRI tasks that test sensory modalities instead of consciousness have been used to show retained brain function even in DoC patients that do not respond to the tennis task. In this work the tennis task and a battery of other tasks including tactile, visual and auditory stimulation, were studied on a group of DoC patients. It was found that none of the patients responded to the task of imagining playing tennis, but retained sensory function could be identified in three out of seven patients. This highlights a strength of fMRI, namely that it can identify retained brain function below the level that is necessary for consciousness. However, the results also show that more than half of the patients studied here, did not show retained brain activation during the fMRI scan. For any of the patients that did not show a response, this can be due to an actual lack of retained brain function, but it can also be due to limitations of the task-based fMRI analysis. The fMRI tasks only test one sensory function at a time, for a short time. Thus a visual fMRI task for example, can only provide information about areas of the brain, that are involved in visual processing. And when vigilance is fluctuating, retained brain function can remain undetected, if vigilance is low during the scan. Functional connectivity analysis is a method to study internal connections between brain areas that is not dependent on an external task. It models the brain as a network of interconnected regions and studies the characteristics of this network. Graph theory is a mathematical field that has found application on many other fields using network analysis, like social sciences, metabolic network modelling or gene network modelling. In fMRI analysis, graph theory has been used to study different phenomena and pathologies and global network properties have been shown reproducibly. The work presented here aims to develop new methods based in graph theory aiding the identification of residual brain integrity. To allow an assessment of the brain network topology and its use in the assessment of residual brain integrity, a novel method was designed based on a graph theoretical measure. The method, termed Cortical Hubs And Related network Topology (CHART) is a two stage procedure. The rst stage identifies statistically significant differences in functional connectivity between two groups, using a measure of the average connectivity of each voxel, the weighted global connectivity. The second stage highlights the topology of the networks associated with those differences. Two fMRI datasets, with different underlying tasks and pathologies were used to test the CHART method. The first dataset was acquired from a group of patients with severe depression. It contrasted the state of the brain before and after successful treatment with electroconvulsive therapy. In this patient group the CHART method was able to identify an area of hyperconnectivity in the depressed state, compared to the treated state. This area of hyperconnectivity was connected to areas that had priorly been shown to be overly connected in the depressed state. The second dataset consisted of DoC patients, that had been extensively assessed behaviourally. Half of the patients were diagnosed to be in a vegetative state, the other half was diagnosed to be in a minimally conscious state. The first stage of CHART identified several areas of difference based on the weighted global connectivity. The second stage highlighted that the observed global differences were due to an overall lack of extended functional connectivity in the vegetative state patients. The addition of a healthy control group in stage two allowed comparison not only between the two DoC groups, but also with the healthy group. In summary it was observed that the spatial extent of the connectivity seen in the minimally conscious group resembles the spatial extent of the connectivity seen in the healthy control group, while the spatial extent of connectivity observed in the vegetative state group was minimal, compared to both healthy and minimally conscious group. Thus the spatial extent of connectivity is a distinguishing property for the vegetative state group studied here. However the first stage of the CHART method is a group based method. In disorders of consciousness, where the underlying pathology is different from case to case, this concept is problematic. Finding a meaningful group of interest is difficult or impossible, because lesions differ in location and extent. Individual differences in connectivity can be expected to be large, and a generalisation of the CHART result might not lead to improved diagnosis for every patient. For diagnosis, the patients individual characteristics must be taken into account. An additional objective of this work was therefore to develop a method to compare a single patient to a group of controls. An approach based on regression modelling was tested but failed to provide the necessary statistical sensitivity to detect impaired connectivity. In conclusion the CHART method developed in this work provides insights into the functional connectivity of a group of DoC patients. To assist diagnosis, further development of a method to compare a single subject to a group of controls will be important.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:606447 |
| Date | January 2013 |
| Creators | Merz, Susanne |
| Publisher | University of Aberdeen |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://digitool.abdn.ac.uk:80/webclient/DeliveryManager?pid=211209 |
Page generated in 0.0031 seconds