大腦由上至下感官預測力與學步兒語言發展之fNIRS研究

Abstract

Emberson 等人2015年發現正常發展的嬰兒其大腦中進行由上而下跨感官訊息預測的能力在6月齡就已到位，該實驗讓6月齡嬰兒先進行聽覺與視覺刺激材料的配對學習，一段時間後，不預期省略原本應該與聽覺刺激一起出現的視覺圖片(視覺缺漏階段)，結果發現6月齡嬰兒在進行視覺缺漏階段時，大腦視覺區仍出現活化現象，因此推論6月齡嬰兒能夠在短時間內習得感官間的配對關係，並且據此對未來訊息進行預測。而最近以臺灣嬰兒為研究對象，採用類似作業典範，進一步發現此彰顯在神經層次的由上而下感官預測力能夠有效預測嬰兒在12個月大時的手勢溝通和詞彙表達能力(Wang et al., 2020）。 有鑑於先前研究對象為年齡較小的受試族群，僅能測量到詞彙層面的能力，無法觀察到大腦的神經預測能力是否也與句子層次的處理能力有關聯。基於此，本研究採用相同的實驗典範以功能性近紅外光譜收集大腦活動資料，但以語言使用更為複雜的16至20月齡學步兒作為研究對象，探究由上而下感官預測與語言發展之關係，尤其是句子層次的發展，是否具有同時相關性。 研究結果發現，16至20月齡學步兒在視覺缺漏的實驗階段，與過去6月齡及12月齡嬰兒有著相同實驗結果，且大腦在視覺缺漏階段於枕葉的血氧濃度活化程度與利用華語嬰幼兒溝通發展量表(幼兒版)所測量到的詞彙表達能力以及句子使用能力(如何使用語言、語言表達複雜度)均達顯著正相關，這個發現為學步兒語言發展與大腦神經由上而下的預測力之關聯性提供了更進一步的支持證據。

Emberson et al.’s (2015) showed that the neural response related to cross-modal top-down sensory predictions is already in place in typically developing 6-month-olds. In their experiment, 6-month-old infants were presented with a visual stimulus soon after they heard a sound such that a sound predicted the presentation of a visual stimulus. After learning this cross-modal association, infants were then presented with trials in which the sound was presented but the predicted visual stimulus was unexpectedly omitted (visual omission trials). Results showed that a robust occipital response was present on visual-omission trials, suggesting that 6-month-old infants can learn the paired relationship in a short time and predict upcoming information. Recently, Wang et al. (2020) adapted the similar experimental procedure and found that the neural response related to cross-modal top-down sensory predictions measured at 6 months was a reliable longitudinal predictor of children’s gesture productions and expressive vocabulary at 12 months. Based on the findings of Wang et al., the present study took a step further to see whether this link can be also identified in older children with more complex language skills. To achieve this aim, a group of toddlers aged between 16 to 20 months were recruited. Functional near-infrared spectroscopy (fNIRS) was used to measure the neural response related to cross-modal top-down predictions using the same fNIRS task as used in the study of Wang et al.. The results revealed positive links between fNIRS predictive signals and toddler’s vocabulary production and syntactic complexity (how to use language, grammatical complexity) measured by the Toddler form of Mandarin-Chinese Communicative Development Inventory. These findings provide additional supports for the hypothesis that the top-down prediction ability manifested at the neural level plays a role in young children’s language development.