| dc.description.abstract | To predict the risk-benefit balance of elective epilepsy surgery as accurately as possible, it is essential that the clinical protocol is systematic, reliable and validated. In clinical practice, the procedures and task selection of language mapping have unfortunately never been standardized. It was expected that clinicians often develop their own protocols by selecting tasks ad-hoc and based on implicit thoughts. Therefore, the aim of this research was to establish a structured, systematic, hypothesis testing approach, based on a language model, in order to achieve a structured rationalized task selection during language mapping. This retrospective explorative intra-individual study was an in-depth case analysis of three characteristic patients intended for elective neurosurgery to treat epilepsy. Data consisted of multiple-day, 24-hour-per-day video registrations. Direct electrical stimulation was used to discrete cortical areas to produce a reversible lesion, presumably simulating the effect of resection of the stimulated site. This study examined the stimulated locations, the performed tasks, the task performances and the clinician's decisions. Results showed that incomplete reports lead to an unstructured and inconsistent working method, causing electrodes and tasks to be over-stimulated and especially under-stimulated based on the 3-5 paradigm. Given the tasks performed, the clinicians focused more on auditory comprehension and spoken production than on written comprehension and production. As a result, not all the language components were examined and could not be distinguished. New tasks were advised and included in two newly constructed flowcharts which serve as a structured rationalized hypothesis-oriented task selection during language mapping. | |
