Auditory change detection has been associated with mismatch negativity (MMN), an event-related potential (ERP) occurring at 100-250 ms after the onset of an acoustic change. Yet, single-unit recordings in animals suggest much faster novelty-specific responses in the auditory system. To investigate change detection in a corresponding early time range in humans, we measured the Middle Latency Response (MLR) and MMN during a controlled frequency oddball paradigm. In addition to MMN, an early effect of change detection was observed at about 40 ms after change onset reflected in an enhancement of the Nb component of the MLR. Both MMN and the Nb effect were shown to be free from confounding influences such as differences in refractoriness. This finding implies that early change detection processes exist in humans upstream of MMN generation, which supports the emerging view of a hierarchical organization of change detection expanding along multiple levels of the auditory pathway.

Auditory deviance detection has been associated with a human auditory-evoked potential (AEP), the mismatch negativity, generated in the auditory cortex 100 -200 ms from sound change onset. Yet, single-unit recordings in animals suggest much earlier (ϳ20 -40 ms), and anatomically lower (i.e., thalamus and midbrain) deviance detection. In humans, recordings of the scalp middle-latency AEPs have confirmed early (ϳ30 -40 ms) deviance detection. However, involvement of the human auditory brainstem in deviance detection has not yet been demonstrated. Here we recorded the auditory brainstem frequency-following response (FFR) to consonant-vowel stimuli (/ba/, /wa/) in young adults, with stimuli arranged in oddball and reversed oddball blocks (deviant probability, p ϭ 0.2), allowing for the comparison of FFRs to the same physical stimuli presented in different contextual roles. Whereas no effect was observed for the /wa/ syllable, we found for the /ba/ syllable a reduction in the brainstem FFR to deviant stimuli compared with standard ones and to similar stimuli arranged in a control block, with five equiprobable, rarely occurring sounds. These findings demonstrate that the human auditory brainstem is able to encode regularities in the recent auditory past to detect novel events, and confirm the multiple anatomical and temporal scales of human deviance detection.

Single neurons in the primary auditory cortex of the cat show faster adaptation time constants to short- than long-term stimulus history. This ability to encode the complex past auditory stimulation in multiple time scales would enable the auditory system to generate expectations of the incoming stimuli. Here, we tested whether large neural populations exhibit this ability as well, by recording human auditory evoked potentials (AEP) to pure tones in a sequence embedding short- and long-term aspects of stimulus history. Our results yielded dynamic amplitude modulations of the P2 AEP to stimulus repetition spanning from milliseconds to tens of seconds concurrently, as well as amplitude modulations of the mismatch negativity AEP to regularity violations. A simple linear model of expectancy accounting for both short- and long-term stimulus history described our results, paralleling the behavior of neurons in the primary auditory cortex.

The ability to detect unexpected novel stimuli is crucial for survival, as it might urge a prompt adaptive response. Human auditory novelty detection has been associated to the mismatch negativity long-latency auditory-evoked potential, peaking at 100-200 ms. Yet, recent animal studies showing novelty responses at a very short latency (about 20-30 ms) in individual neurons already at the level of the midbrain and thalamus suggest that novelty detection might be a basic principle of the functional organization of the auditory system, expanding from lower levels in the brainstem along the auditory pathway up to higher-order areas of the cerebral cortex. To test this suggestion, we here measured auditory brainstem and middle latency response (MLR) to frequency novel stimuli embedded in an oddball sequence. To oversee refractoriness confounds a 'control block' was used. The results showed that occasional changes in auditory frequency information were detected as early as 30 ms (Pa waveform of the MLR) after stimulus onset. The control block precluded these effects as resulting merely from refractoriness, altogether supporting the notion of 'true' early auditory change detection in humans, matching the latency range of auditory novelty responses described in individual neurons of subhuman species. Our results suggest that auditory change detection of frequency information is a multistage process that occurs at the primary auditory cortex and is transmitted to the higher levels of the auditory pathway.

Prompt detection of unexpected changes in the sensory environment is critical for survival. In the auditory domain, the occurrence of a rare stimulus triggers a cascade of neurophysiological events spanning over multiple timescales. Besides the role of the mismatch negativity (MMN), whose cortical generators are located in supratemporal areas, cumulative evidence suggests that violations of auditory regularities can be detected earlier and lower in the auditory hierarchy. Recent human scalp recordings have shown signatures of auditory mismatch responses at shorter latencies than those of the MMN. Moreover, animal single-unit recordings have demonstrated that rare stimulus changes cause a release from stimulusspecific adaptation in neurons of the primary auditory cortex, the medial geniculate body (MGB), and the inferior colliculus (IC). Although these data suggest that change detection is a pervasive property of the auditory system which may reside upstream cortical sites, direct evidence for the involvement of subcortical stages in the human auditory novelty system is lacking. Using eventrelated functional magnetic resonance imaging during a frequency oddball paradigm, we here report that auditory deviance detection occurs in the MGB and the IC of healthy human participants. By implementing a random condition controlling for neural refractoriness effects, we show that auditory change detection in these subcortical stations involves the encoding of statistical regularities from the acoustic input. These results provide the first direct evidence of the existence of multiple mismatch detectors nested at different levels along the human ascending auditory pathway.