Back

Perception Of Non-Verbal Prosody In Children With Asd

Martynova, O.; Portnova, G.; Maslennikova, A.

2020-12-07 neuroscience
10.1101/2020.12.07.414201 bioRxiv
Show abstract

Paralinguistic features of the speaker, such as prosody, temp, loudness, and dynamics, are an important marker of a persons emotional state. The deficit of processing of emotional prosody could be preferably associated with the impairments in individuals with ASDs social behavior. The following two groups of children participated in our study: 30 preschoolers from 4 to 6 years old in the target group (39.1 {+/-} 6.4 scores by Childhood Autism Rating Scale), 24 preschoolers of the control group from 4 to 6 years in the control group. The prosody stimuli were the combination of syllables, said with intonations of "joy," "angry," "sadness," "fear," and "calm." Fast Fourier transform (FFT) is used to analyze power spectrum density (PSD). The resulting normalized spectrum was integrated over unit width intervals in the range of interest (2 to 20 Hz with a step in 1Hz). Children with ASD, similarly to TD children, showed the most pronounced differenced of EEG in response to prosodics of fear and anger. The significant groups differences in PSD were detected for sad and joy intonations. Indexes of EEG differences between pleasure and painful intonations were significantly higher in the control group than children with ASD and between sadness and calm or joy and calm intonations. This paper makes up two main contributions: In general, we obtained that children with ASD have less response to a human voices emotional intonation. The physical characteristics of stimuli are more critical than a sign of emotions. The effect of EEG spectral power has hemisphere specialization in the healthy control group, but not in ASD children. Since spectral power for negative emotions in the target group is higher, we proposed that ASD children worse recognize positive emotions than negative emotions.

Matching journals

The top 12 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.