Duville Mathilde Marie, Alonso-Valerdi Luz Maria, Ibarra-Zarate David I
Neuroengineering and Neuroacoustics Research Group, Tecnologico de Monterrey, Escuela de Ingeniería y Ciencias, Monterrey, Mexico.
Front Hum Neurosci. 2021 Feb 26;15:626146. doi: 10.3389/fnhum.2021.626146. eCollection 2021.
Socio-emotional impairments are key symptoms of Autism Spectrum Disorders. This work proposes to analyze the neuronal activity related to the discrimination of emotional prosodies in autistic children (aged 9 to 11-year-old) as follows. Firstly, a database for single words uttered in Mexican Spanish by males, females, and children will be created. Then, optimal acoustic features for emotion characterization will be extracted, followed of a cubic kernel function Support Vector Machine (SVM) in order to validate the speech corpus. As a result, human-specific acoustic properties of emotional voice signals will be identified. Secondly, those identified acoustic properties will be modified to synthesize the recorded human emotional voices. Thirdly, both human and synthesized utterances will be used to study the electroencephalographic correlate of affective prosody processing in typically developed and autistic children. Finally, and on the basis of the outcomes, synthesized voice-enhanced environments will be created to develop an intervention based on social-robot and Social Story for autistic children to improve affective prosodies discrimination. This protocol has been registered at BioMed Central under the following number: ISRCTN18117434.
社会情感障碍是自闭症谱系障碍的关键症状。本研究计划如下分析与自闭症儿童（9至11岁）情感韵律辨别相关的神经元活动。首先，将创建一个包含男性、女性和儿童用墨西哥西班牙语说出的单个单词的数据库。然后，提取用于情感特征描述的最佳声学特征，接着使用立方核函数支持向量机（SVM）来验证语音语料库。结果，将识别出情感语音信号的人类特异性声学特性。其次，将对那些识别出的声学特性进行修改，以合成录制的人类情感语音。第三，人类语音和合成语音都将用于研究正常发育儿童和自闭症儿童情感韵律处理的脑电图相关性。最后，根据研究结果，将创建合成语音增强环境，以开发基于社交机器人和社交故事的干预措施，帮助自闭症儿童提高情感韵律辨别能力。该方案已在BioMed Central注册，注册号为：ISRCTN18117434。
