Presentation
Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Lightning Talks
Brain mechanisms of statistical sequence learning in children with DLD – methods and initial results of fMRI experiments.
Poster B66 in Poster Session B, Tuesday, October 24, 3:30 - 5:15 pm CEST, Espace Vieux-Port
This poster is part of the Sandbox Series.
Martyna Bryłka1, Hanna Cygan1, Jakub Wojciechowski1, Tomasz Wolak1; 1The Institute of Physiology and Pathology of Hearing
The brain is known to rapidly find structure and meaning in unfamiliar streams of stimuli, even with minimal external feedback. One type of probabilistic information is statistical sequence and the most prominent example of natural sequential information is linguistic code. Statistical sequence learning is considered as a sub-process of procedural learning, which was proposed as a one of the core functions underlying language acquisition. Subsequently, its dysfunctions became an important candidate for a mechanism underlying core symptoms of Developmental Language Disorder (DLD). The neural mechanisms of statistical learning (SL) involve basal ganglia, frontal lobes, and medial temporal cortex. Importantly, the involvement of specified networks is dependent on the stage of the learning process and the level of statistical complexity. There are limited findings on neural mechanisms of SL in DLD. One study revealed that adults with DLD activate a typical neural network, however, a regional analysis indicated hyperactivation of regions associated with language processing (Plante et al., 2017). Present project aims to verify hypothesis: - The SL impairment in DLD is related to atypical brain circuit including basal ganglia, frontal and temporal lobes; - The level of SL impairment varies depending on the stimulus type (verbalizable vs. non-verbalizable) and is interdependent with the underlying engagement of specific brain structures. We examined 39 children with DLD (age 7-9) and 46 TD children matched in terms of age and non-verbal IQ with no comorbid dysfunction. All were diagnosed with language, IQ battery and ASD screening. We developed the fMRI paradigm for children of sequential SL that was based on Wang's et al. (2017) paradigm. In contrast to the classic SRT, our task does not involve visuo-spatial information and is independent of motor performance. We used two types of stimuli (verbalizable vs non-verbalizable): 4 abstract symbols, and 4 pictures of recognizable animals. We included structured sequences of 1st-order (Markov’s model): probability of one symbol depends on the preceding symbol. Random sequences were a control condition. During fMRI, after being exposed to the sequence, participants were to decide which of the symbols is going to appear next. The distribution of responses was rated as a behavioral result. Children underwent two fMRI sessions separated with one-week training including 4-sessions of SL task. Preliminary results from children will be included on the poster presentation to discuss. In the fMRI experiment conducted initially to validate the SL-procedure with 36 TD adults (age 20-45) we obtained differences for statistical-random sequences in the basal ganglia: Left Putamen for all stimuli; Left Caudate for verbalizable stimuli; Right Putamen for non-verbalizabe stimuli. Moreover for verbalizable statistical sequences, there was an activity increase in the Right Insula and Left MFG. Contrast of random-statistical revealed effects in: Left Caudate, Left Nucleus-Accumbens and Right Putamen for non-verbalizable stimuli; Additionally, processing of random sequences showed an increase in activity in the right inferior frontal cortex. In conclusion, neural processing of statistical sequence is specific for stimulus type. We expect effects of DLD vs TD comparison around similar structures, that will show specificity for stimulus type.
Topic Areas: Disorders: Developmental, Language Development/Acquisition