| 摘要: | 在日常生活中，嘈雜環境常常造成語句辨識困難，對聽力損失者與高齡族
群尤為明顯。多感官線索（如:嘴型與聲音）的整合可以提升在嘈雜環境中語句
辨識的效益。文獻指出，音樂家通常展現較佳的多感官線索整合能力，例如:他
們能更精準的將聽覺與視覺的刺激整-合，並且擁有較窄的時間整合視窗
（temporal binding window）。然而，大多數相關研究只比較音樂家和非音樂
家，較少考量到音樂家之間的個體差異。其中，絕對音感（Absolute Pitch,
AP）能力與自閉症特質中的多感官整合處理模式可能有關聯性。因此，本研究
探討絕對音感這個先天音樂能力，如何影響視聽覺線索的整合，進而影響在嘈
雜環境中語句辨識之行為表現與腦神經機制。
在行為實驗中，74 名受試者依絕對音感能力分為 3 組:具絕對音感的音樂家
組（AP）、非絕對音感的音樂家組（non-AP）以及非音樂家組（nonmusician），受試者參與視聽同步性判斷和時間順序判斷實驗，實驗操弄了以
下 13 種刺激起始不同步（SOAs: ±360, ±300, ±240, ±180, ±120, ±60, 和 0 毫秒）
的中文語句，以量測其時間整合視窗。為了評估嘈雜環境中的視覺和聽覺整合
的語句理解效益，我們分析了 3 組受試者在四種訊噪比（0、−6、−9、−12
dB）的背景雜訊中，分別在單純聽覺（Audio-only, AO）、單純視覺（Visualonly, VO）以及視聽覺條件（Audiovisual, AV）三種情況的語句辨識表現。實驗
亦測試與絕對音感相關的變項，包含音樂能力以及自閉症特質。結果顯示，具
有絕對音感的音樂家展現出較寬的時間整合視窗，代表其整合視聽覺刺激的能
力較弱，且此能力與較高程度的自閉症特質呈正相關。此外，較差的整合能力
和在嘈雜環境中辨識語句所獲得的視聽覺效益較低有關，顯示絕對音感可能為調節多感官整合以及嘈雜中的語言理解能力的其中一個關鍵因素。
在功能性磁振造影（fMRI）實驗中則測試絕對音感音樂家、非絕對音感音
樂家以非音樂家組在不同視聽覺條件下處理語句的大腦活化差異。受試者聆聽
在三種訊噪比（無噪音、0、−12 dB）下所呈現之單純聽覺（AO）、單純視覺
（VO）, 或視聽覺條件（AV）的語句，並以按鍵回答有出現的語詞。結果顯
示，相較於非絕對音感者，具有絕對音感的音樂家在單一感官（純聽覺、純視
覺）的條件下，其顳上迴（superior temporal gyrus, STG）已出現與感官整合相
關的活化。在視聽覺條件（AV）下，顳上迴與額上迴（superior frontal gyrus,
SFG）區域在絕對音感組有更高的大腦活化，這些腦區與多感官處理有密切相
關。此外，顳上迴的活化程度與受試者的自閉症特質呈現正相關。過去文獻指
出具高自閉症特質者在視聽整合上，顳葉和額葉區域活化較高。這或許顯示絕
對音感的音樂家可能在多感官整合腦區的活化模式上，與具有較高自閉症特質
的個體有所相似。
研究結果顯示絕對音感為調節多感官語句辨識表現與神經機制的一種先天
音樂能力。相較於過去文獻顯示音樂家具有較佳的多感官線索整合能力，本研
究指出具絕對音感者展現出較弱的多感官整合能力，在嘈雜環境中辨識語句
時，從語句之多感官線索得到的效益較少。此外，在處理視聽線索時，絕對音
感者與較高的自閉症特質呈現正相關，這或許暗示他們在感官整合的神經機制
上可能存在某些相似性。未來期能應用於發展結合多感官線索的輔助，以提升
日常溝通中特別是在嘈雜環境中的語句理解能力。;Integrating across multiple senses is associated with benefits for speech
processing, particularly in challenging listening environments, particularly for
individuals with hearing loss and older adults. Multisensory integration, such as
combining visual and auditory cues, can improve speech perception under noisy
conditions. Musicians have been shown to possess enhanced multisensory integration
abilities, often reflected a narrower temporal binding window (TBWs). However,
previous studies often compare musicians and non-musicians, how individual
differences among musicians remains unclear. This study investigated how absolute
pitch (AP), a genetic musical trait among musicians, influences the perceptual and
neural mechanisms underlying musicians′ multisensory integration and audiovisual
benefits for speech perception in noise.
In the behavioral experiments, 74 musicians and non-musicians divided into AP
musician, non-AP musician, and non-musician groups completed the audiovisual
synchrony and temporal order judgement tasks under varying stimulus onset
asynchronies (SOAs: ±360, ±300, ±240, ±180, ±120, ±60 and 0 ms) of spoken
Mandarin sentences, to assess their TBW. To examine how participants integrate
visual and auditory cues to support speech comprehension in noisy environments,
they completed the audiovisual speech-in-noise task under various signal-to-noise
ratios (SNRs: 0, –6, –9, –12 dB) under three sensory conditions: audio-only (AO),
visual-only (VO), and audiovisual (AV). Additional measures included musical
aptitude and autism spectrum traits. Behavioral results revealed that AP musicians
exhibited significantly wider TBWs, indicating weaker audiovisual integration ability, which was positively correlated with higher levels of autism spectrum traits.
Additionally, this reduced temporal sensitivity was linked to smaller audiovisual
benefits for speech perception in noise, suggesting that integration ability may
influence higher-level functions such as speech comprehension.
In functional magnetic resonance imaging (fMRI) experiment, participants were
examined for differences in brain activation during sentence processing under audioonly (AO), visual-only (VO), and audiovisual (AV) speech perception in noise at
different SNRs (no noise, 0, –12 dB). Participants completed the speech recognition
task in noise, responding by pressing one of four buttons on a response pad during
fMRI scanning. Compared to non-AP musicians, AP musicians showed increased
activation in the superior temporal gyrus (STG) during unimodal (AO and VO)
conditions, regions typically associated with multisensory integration. Under AV
conditions, higher activation was observed in both the STG and the superior frontal
gyrus (SFG) among the AP group. Notably, STG activity positively correlated with
participants′ autistic traits. These findings consistent with previous research showing
enhance temporal and frontal lobe activation during audiovisual integration in
individuals with higher autistic traits, suggesting a potential link neural mechanisms
between AP and autistic traits.
Overall, the results highlight AP as an innate musical ability that modulates both
behavioral and neural mechanisms of multisensory speech perception. In contrast to
previous findings suggesting superior multisensory integration in musicians, our
results indicate that AP musicians exhibited significantly weaker audiovisual
integration and derived less benefit from audiovisual cues in noisy environments. In
addition, AP is linked to higher autistic traits, and autism has been associated with impaired multisensory integration. These findings may inform the development of
multisensory cue-based interventions to improve speech comprehension, particularly
in challenging listening conditions. |
