Work

Assessment of self and environmental consciousness in unresponsive patients: Detecting complete-locked-in syndrome wit brain-computer interfaces

Year: 2024
Type: dissertation
Abstract: When I was in my first internship as a medical resident, I took care of a patient few monthsafter his brainstem stroke. He had open eyes, but no communication code. He remainedmotionless in his bed, only his gaze shifted a little. At certain times, his eyes seemed to followus, and at others, to turn away. But maybe that was just my interpretation. When I went intohis room, I was plagued with questions. Was he in pain? What was his mood? Was I botheringhim? Or was he happy to have some company? Was I right to infer consciousness from thoseopen eyes? How to prove consciousness in the absence of communication?Each of the many questions we ask ourselves when faced with an unresponsive person couldbe the starting point for a thesis. One way of starting to answer some of these questions is tocollect retrospective reports from patients who have emerged from a non-responsive state.This is the case for most persons presenting with a locked-in syndrome (LIS), which is a globalparalysis due to a brainstem injury. Most persons in LIS have experienced a coma in the acutephase, followed by a period where they could open their eyes, but with no meaningful way tocommunicate yet. It was the case with the patient I was confronted with during my firstresidency internship. It is not clear if this period is a Complete LIS (with awareness andpreserved cognition without any voluntary motor control) or a state of disorder ofconsciousness. We designed a questionnaire dedicated to persons in the chronic phase of LIS.Thanks to the French association for the locked-in syndrome (ALIS), 41 patients answered thequestionnaire. The results uncovered that some of them did experience a phase of completelocked-in state. These results are exposed in the first study of this thesis.New tools claim to be able to re-establish a form of communication, even in the completeabsence of the ability to move: brain-computer interfaces (BCIs). They aim at extractinginformation directly from the brain. We decided to assess such a BCI to try to help thesepatients. Previous studies had demonstrated that there is a high prevalence of visualimpairments in these patients, preventing the use for visual BCI. We then chose to test anauditory one. It is obviously complicated to test a new tool with patients who are unresponsive,and we don’t know whether they are aware or not. As we shall see in this thesis, the profile ofunresponsive patients is very heterogeneous. So, we thought it would be simpler to start testingand adjusting the interface with patients in a locked-in state, but having a residualcommunication code.The second study describes this pilot trial and its results. We included 7 patients withtetraplegia, anarthria and facial paralysis. Three had a brainstem injury and 4 an AmyotrophicLateral Sclerosis (ALS), a degenerative disease of the motor neurons. We realized that thesepatients had troubles to control the BCI, and when they did, it was with differentelectrophysiological signals from those of healthy subjects.The locked-in syndrome is an extreme condition where an injury in the anterior part of the ponsinterrupts the pyramidal tracts descending to the motor neurons innervating the limbs, thetrunk and the face. The pyramidal tract is classically supposed to be "an exit route" conveyingthe motor execution information. Hence, the impact of its injury is usually considered as amatter of disconnection between the motor cortex and the muscles. It is not supposed toimpact cognition. De facto, these patients can usually manage complex means ofcommunication, like letter boards, requiring an efficient working memory and strong languagePage 4 sur 139abilities. But then, how can we explain that brain-machine interfaces struggle to detect signs ofvoluntary modulation of brain activity?Our third study covers and synthesizes a large literature, from clinical studies to cognitiveneuroscience and theoretical ones, and shows the complexity of the links between attentionand motor control. This paper questions the possibility to have preserved attention when allthe corticospinal and corticonuclear pathways are altered. Specifically, I propose thatattentional processes are likely altered in conjunction with, or as a result of, motor disorders.This hypothesis aligns with the premotor theory of attention, which is supported by bothbehavioral and neurophysiological evidence. A consequence of this hypothesis, is that BCIcontrol could be impeded in most patients with global paralysis, as BCI control requires strongattentional abilities. Beyond proposing this hypothesis, this paper also challenges our BCIcommunity to further explore the clinical causes of BCI failures.Building on these data, we decided to test the hypothesis that the presence of an oculomotorimpairment is correlated with an impairment of voluntary attentional control. We designed andpre-registered a study assessing jointly oculomotor control and three active EEG paradigmswith 30 healthy subjects and 54 patients. The fourth study presents this main experimentalmulticenter experimentation and its preliminary results.While it is frustrating to see the many obstacles encountered when trying to overcome globalparalysis, we also highlight how essential it is to clearly define these obstacles in order to bettercircumvent them. It is often the technological obstacles that are the subject of discussion, andof development prospects. Physiopathology remains in the background. We argue thatphysiopathology is a facet of the translation pitfalls encountered by BCIs and should be bettertaken into account to optimize future developments. Some obstacles are first of all a lack ofknowledge of the end users of BCIs. It is of course very difficult to get to know users whosecommunication is absent or restricted due to neurological damage. However, we show herethat there are now different ways to start inferring the cognitive functions of these patientsand their needs, from clinical examination to complex neuro-imaging assessment. Based on thisnew multimodal semiology of unresponsive patients and those with locked-in syndrome, wepropose research perspectives to better understand the impact of paralysis, and to design BCIsin a more individualized manner. We are also exploring avenues to improve the disseminationof the results of BCI studies, to allow patients and the population access to clear information.
Source: HAL
Language: English
Field: Medicine
Open Access status: green
OpenAlex ID: W6998632247
Cites: 315
Cited by:
Related to: 9