Temporal Coding of Voice Pitch Contours in Mandarin Tones

Area of Science:

• Neuroscience
• Auditory Neuroscience
• Speech Processing

Background:

• Accurate perception of time-variant pitch is crucial for speech recognition, especially in tonal languages like Mandarin.
• Previous research indicates auditory nerve and cochlear nucleus encode pitch via phase-locked neural activity.
• The encoding mechanisms of time-variant periodicity pitch in the inferior colliculus (IC) remain largely unknown.

Purpose of the Study:

• To investigate how the inferior colliculus (IC) encodes the time-variant periodicity pitch of natural speech.
• To compare the encoding capabilities of local field potentials (LFPs) and single neuron activity within the IC for pitch information.

Main Methods:

• Simultaneous recording of LFPs and single neuron activity from the IC of urethane-anesthetized guinea pigs.
• Stimuli included Mandarin syllables with four distinct lexical tones (flat, rising, falling-rising, falling).
• Analysis focused on temporal information, pitch strength using autocorrelograms, and neural synchronization to the fundamental frequency.

Main Results:

• 93% of LFPs demonstrated robust encoding of periodicity pitch.
• LFP pitch strength was significantly higher for rising tones compared to flat and falling tones (p < 0.001).
• 47% of single neurons showed significant synchronization to the stimulus's fundamental frequency, indicating robust temporal coding.

Conclusions:

• The IC effectively encodes time-variant speech pitch, with LFPs showing a more widespread representation than single neurons.
• Findings suggest a neural representation transition within the IC, moving from direct temporal coding in spike trains to other neural representations.
• This study provides insights into the neural basis of pitch perception essential for understanding tonal languages.