Eliciting verb inflection in the English language - The Verb and Noun Test (VAN) for Presurgical Language Mapping with navigated TMS and Intraoperative DES

Eliciting Verb Inflection in the English Language – The Verb and Noun Test (VAN) for Presurgical Language Mapping with navigated TMS and Interaoperative DES Ann-Katrin Ohlerth1, Antonio Valentin2, Keyoumars Ashkan3, Francesco Vergani3, Molood Sadat Safavi4, Frank Zanow4 & Roelien Bastiaanse1,5 1Center for Language and Cognition Groningen, the Netherlands 2Institute of Psychiatry, Psychology & Neuroscience (IoPPN), Kings College London, UK 3Neurosurgical Department, King's College Hospital, London (UK) 4eemagine Medical Imaging Solutions GmbH, Berlin, Germany 5National Research University, Higher School of Economics, Moscow, Russian Federation Introduction: To ensure maximal removal of tumour in eloquent areas of the brain while preserving function, Direct Electrical Stimulation (DES) during awake brain surgery is the gold standard for intraoperative language mapping. To reduce the duration of the surgery and, thus, the stress for the patient, preoperative language mapping using navigated Transcranial Magnetic Stimulation (nTMS) can be beneficial (Tarapore, 2016). The method is, however, still not widely used in clinical practice. Reason for this could be the low predictive value of nTMS compared to DES mapping (Picht et al., 2013). In the recent years, advances have been made to further improve preoperative language mapping under nTMS. Parameters such as intensity, frequency, duration etc. of the stimulation, picture-presentation-time, inter-picture-interval of the stimuli etc. have been tested out and consensus has been reached that is used by the majority of centres (Krieg et al., 2017). However, the question of choice of task has not been further explored: whereas in the intraoperative situation, a variety of tasks is used and carefully chosen based on the individual’s characteristics, most centres only rely on one task, namely object naming, for preoperative mapping of language. It is known from stroke-induced aphasia research, that tasks involving only noun retrieval and production can hardly capture the individual’s language skills. The necessity of including verb tasks is evident (Bastiaanse, Wieling & Wolthuis, 2016). While some studies (Hauck et al., 2015, Hernandez-Pavon et al., 2014) have looked into the comparison of noun vs. verb tasks under nTMS, they neglected the actual contribution a verb task can have: the elicitation of the inflected form of a verb. However, while naming action in English, participants tend to use the progressive form of the verb (he is sleeping). This would diminish the morphological processes required for inflection, compared to inflection in tenses such as simple present and past (he sleeps/he slept). Therefore, adding a lead-in phrase with an adverb to the picture can help triggering tenses other than simple progressive (Daily, he sleeps). This has not been tested before regarding nTMS mapping. In this validation study, we established a reliable item list for both object and inflected action naming tasks for British English (BE) that uses lead-in phrases triggering one-word answers and that is feasible to use under nTMS parameters. We present a validated picture set with each 50 items for object naming and 50 for action naming with a high naming agreement. Methods: From a collection of around 400 available black-and-white drawings (Bastiaanse et al., 2016), the researchers chose 90 suitable pictures to elicit possible one-word targets in English. This preselection was shown to 5 BE native speakers with the instructions to name the object/action on the picture. Only pictures consistently named with one label were considered good depictions and chosen for the next step. Next, the remaining 80 object and 74 action stimuli were implemented in an experiment, following parameters commonly used in nTMS language mapping. Participants were instructed to name the pictures with the first word coming to their mind in one word. For the object naming task, they were asked to each time complete and read out the sentence “this is a…”, printed above. For the action naming task, elicitation of the progressive form of the verb was avoided, and instead the inflection of the verb in different tenses was triggered. For the first half of the stimuli, the participants were asked to complete and read out the sentence “Daily, he…”, triggering the the verb in simple present. In the second half, the printed lead-in phrase “Yesterday, he…” triggered the verb in simple past. The picture presentation time was 700ms for the objects and 1000ms for the actions with an inter picture interval of 2500ms comparable to conditions under TMS. After two practice items, the test began and was not further interrupted. Answers were recorded. The test took around 12 minutes. 28 native speakers of BE (15 female; age range: 23-65, age average: 40.91 with a wide range of educational background) took part in London, UK. Result Only items that were consistently named by at least 80% of the participants were selected for the final item list. 71 object stimuli and 50 action stimuli proofed to fulfill the criterion. To keep the lists even, another 21 objects were excluded. The final item lists consist of 100 items sorted by ascending values for Age of Acquisition of the words. All items are also matched for frequency, length in syllables, animacy (for objects); regularity, transitivity, number of arguments, instrumentality and name relatedness to a noun (for the actions). In the action naming paradigm, both subtests (present and past) have a balanced number for irregular vs. regular verbs. No difference was observed in the correct/incorrect naming of the irregular vs. regular verbs, neither within nor across tasks. Discussion: This validation study resulted in a reliable set of stimuli: we established two tasks that are capturing the linguistic abilities more adequate than a noun task alone could. The tasks are balanced for age of acquisition and frequency, as well as irregularity for the verbs. Moreover, all stimuli are controlled for linguistic factors, known to be impaired in clinical populations, and can, hence, be sorted participant-tailored. The inclusion of a lead-in phrase was necessary to trigger the inflected form of the verbs. While it makes the visual input more complex and the answer longer, it did not hinder the participants for completing the tasks correctly and in time. Even for the more complex pictures depicting less frequent actions, the picture presentation time has shown to be sufficient to recognize and name the target, yet short enough to be challenging and to stay as close as possible to the commonly used protocols. This proofs the paradigm to be feasible for usage under nTMS. The test will be made freely available for neurosurgical teams; it can be adapted to other languages. References: Bastiaanse, R., Wieling, M., & Wolthuis, N. (2016). The role of frequency in the retrieval of nouns and verbs in aphasia. Aphasiology, 30(11), 1221-1239. Krieg, S. M., Lioumis, P., Makela, J. P., Wilenius, J., Karhu, J., Hannula, H., ... & Islam, M. (2017). Protocol for motor and language mapping by navigated TMS in patients and healthy volunteers; workshop report. Acta neurochirurgica, 159(7), 1187-1195. Hauck, T., Tanigawa, N., Probst, M., Wohlschlaeger, A., Ille, S., Sollmann, N., ... & Krieg, S. M. (2015). Task type affects location of language-positive cortical regions by repetitive navigated transcranial magnetic stimulation mapping. PLoS One, 10(4), e0125298. Hernandez-Pavon, J. C., Makela, N., Lehtinen, H., Lioumis, P., & Makela, J. P. (2014). Effects of navigated TMS on object and action naming. Frontiers in human neuroscience, 8, 660. Rofes A, De Witte E, Marie¨n P, Bastiaanse R (2012) The verb in sentence context test (VISC). Rijksuniversiteit Groningen, Groningen)