Mismatch negativity: clinical and other applications

The perspectives of application of the mismatch negativity (MMN), generated by the brain's automatic response to change in auditory stimulation, are discussed. In light of the fact that the MMN (and its magnetic equivalent MMNm) currently provides the only objective measure of the accuracy of the central auditory function, these perspectives appear very promising. The MMN can be measured in the absence of attention and task requirements, which makes it particularly suitable for testing different clinical populations and infants. Furthermore, the MMN enables one to evaluate the accuracy of auditory discrimination separately for any acoustic feature, such as frequency, intensity and duration, and for learned categories, such as the phonemes of a particular language. In addition, by measuring the decay of the MMN amplitude as a function of the interstimulus interval, it is possible to estimate the duration of sensory (echoic) memory.     

Acute and chronic effects of alcohol on preattentive auditory processing as reflected by mismatch negativity

Chronic alcoholism, a major worldwide health problem, is associated with a variety of neurocognitive changes in the afflicted individuals. The precise neurophysiological basis of these changes is not yet understood. Mismatch negativity (MMN) is a preattentive event-related potential component indexing cortical auditory memory traces and automatic change detection in the brain that can be used to study the neural basis of cognitive impairments in various neurodegenerative diseases. MMN studies have revealed that even a low dose of acute alcohol significantly impairs automatic change detection and involuntary attention shifting. Recent MMN results on chronic alcoholism in turn suggest that auditory sensory traces decay slightly faster and are substantially more vulnerable to the distracting effect of backward masking in alcoholics than in healthy subjects. Furthermore, chronic alcohol abuse might accelerate the age-related impairment of automatic change detection. There is also evidence that the MMN changes might predict impaired performance in behavioral memory and attention tasks in alcoholics. In MMN studies of detoxified alcoholics, however, many confounding factors have to be taken into account. For instance, postwithdrawal brain hyperexcitability might be associated with a slightly enhanced or accelerated MMN/MMNm (the magnetic equivalent of MMN). In sum, MMN and MMNm provide an objective noninvasive tool for exploring the neurophysiological functional deficits related to both acute alcohol intoxication and chronic alcoholism.     

Automatic detection of frequency changes depends on auditory stimulus intensity.

OBJECTIVE: A cortical cognitive auditory evoked potential, mismatch negativity (MMN), reflects automatic discrimination and echoic memory functions of the auditory system. For this study, we examined whether this potential is dependent on the stimulus intensity. DESIGN: The MMN potentials were recorded from 10 subjects with normal hearing using a sine tone of 1000 Hz as the standard stimulus and a sine tone of 1141 Hz as the deviant stimulus, with probabilities of 90% and 10%, respectively. The intensities were 40, 50, 60, 70, and 80 dB HL for both standard and deviant stimuli in separate blocks. RESULTS: Stimulus intensity had a statistically significant effect on the mean amplitude, rise time parameter, and onset latency of the MMN. CONCLUSION: Automatic auditory discrimination seems to be dependent on the sound pressure level of the stimuli.     

Possible neuronal refractory or recovery artifacts associated with recording the mismatch negativity response

The mismatch negativity (MMN) is an event-related potential that involves a negative voltage shift of baseline electroencephalographic (EEG) activity in the approximate latency window of the N1 and P2 cortical potentials in response to new or novel sounds. The MMN is present at birth and has been hypothesized to serve as an automatic preconscious detector of changes in the auditory environment. Research paradigms used to extract the MMN response from EEG activity have a potential problem related to neuronal refractoriness or recovery. Both N1 and P2 are known to increase in amplitude with longer interstimulus intervals (ISIs). The MMN extraction procedures involve mathematical subtraction of waveforms elicited by standard sounds (with short ISIs) from those recorded to rare deviant sounds (with longer ISIs). Any ISI-dependent amplitude changes in N1 and/or P2 could therefore alter the morphology of the resulting difference wave and lead to misinterpretation of the nature of the underlying MMN generators. We tested 12 young females and found that the MMN can be influenced by ISI-dependent refractory effects that may modify the waveform morphology. This has important clinical implications since the MMN is being investigated as an assessment tool.     

Mismatch negativity: different water in the same river

The mismatch negativity (MMN) is a frontal negative deflection in the human event-related potential that typically occurs when a repeating auditory stimulus changes in some manner. The MMN can be elicited by many kinds of stimulus change, varying from simple changes in a single stimulus feature to abstract changes in the relationship between stimuli. The main intracerebral sources for the MMN are located in the auditory cortices of the temporal lobe. Since it occurs whether or not stimuli are being attended, the MMN represents an automatic cerebral process for detecting change. The MMN is clinically helpful in terms of demonstrating disordered sensory processing or disordered memory in groups of patients. Improvements in the techniques for measuring the MMN and in the paradigms for eliciting it will be needed before the MMN can become clinically useful as an objective measurement of such disorders in individual patients.     

Intracranial evidence for differential encoding of frequency and duration discrimination responses

OBJECTIVE: To identify cortical areas involved in auditory detection and discrimination of frequency and duration stimuli in an awake child. DESIGN: Single case study recording intracranial auditory event-related potentials to auditory odd-ball stimuli varying in duration and/or frequency. RESULTS: N1 wave to detection was recorded maximally just above the Sylvian fissure 1 cm posterior to the mismatch negativity (MMN) response to discrimination. Frequency MMN overlapped with the N1 whereas duration MMN appeared 100 msec later. MMN to both duration and frequency appeared as an additive bifid response. CONCLUSIONS: It is suggested that feature-specific neuronal networks are activated after changes in sounds, which may underpin a fast parallel preattentive process (MMN).     

Maturation of the mismatch negativity: effects of profound deafness and cochlear implant use

The use of cochlear implants to restore auditory sensation in deaf children is increasing, with a trend toward earlier implantation. However, little is known about how auditory deprivation and subsequent cochlear implant use affect the maturing human central auditory system. Our previous studies have demonstrated that the obligatory auditory evoked potentials (AEPs) of implanted children are very different from those of normal-hearing children. Unlike the obligatory potentials, which primarily reflect neural responses to stimulus onset, the mismatch negativity (MMN) provides a neurophysiological measure of auditory short-term memory and discrimination processes. The purpose of this investigation is to review our studies of the effects of auditory deprivation due to profound deafness and cochlear implant use on the maturation of the MMN in children, placed in the context of overall age-related changes in the AEPs. The development and application of a statistical technique to assess the MMN in individuals is also reviewed. Results show that although the morphology of the obligatory AEPs is substantially altered by the absence of a normal N(1) peak, the MMN is robustly present in a group of implanted children who have good spoken language perception through their device. Differences exist in the scalp distribution of the MMN between implanted and normal-hearing children. Specifically, the MMN appears to be more symmetrical in amplitude over both hemispheres, whereas it is initially much larger over the contralateral hemisphere in normal-hearing children. These findings suggest that, compared to N(1), the MMN is a better measure of basic auditory processes necessary for the development of spoken language perception skills in profoundly deaf children and adults who use a cochlear implant.     

健康青年人失匹配负波的提取及刺激偏差对波形的影响

目的 研究失匹配负波(mismatch negativity,MMN)波形提取的相关因素及刺激偏差对波形的影响,获得一套实验室标准值.方法 以oddBall序列给予21名健康青年志愿者短纯音刺激,分为频率偏差刺激和强度偏差刺激两组,每组中又包含3个刺激系列,共6个小组.用偏差刺激引出的事件相关电位(event-related potential,ERP)减去标准刺激引出的ERP即得到MMN波形.记录各组潜伏期与波幅并两两比较,分析听刺激偏差程度对MMN的影响.结果 本实验诱发出了健康青年人的MMN波形,并获得了潜伏期与波幅的正常值.频率偏差组中,频率差异为2 kHz时的MMN潜伏期[(155.81±29.08)ms]与频率差异为1 kHz时[(182.89±45.85)ms,(183.32±43.33)ms]相比,明显缩短(P值均<0.05);差异程度相同时,改变标准刺激与偏差刺激频率的绝对值,MMN潜伏期没有明显变化;各小组间MMN的波幅[(3.85±2.22)、(2.90±2.05)、(2.66±2.12)μV]差异无统计学意义(P值均>0.05).强度偏差组中,强度差异为加dB时MMN潜伏期[(157.04±34.87)ms]与强度差异为10 dB时[(184.46±38.05)ms,(186.24±42.36)ms]相比,明显缩短(P值均<0.05);差异程度相同时,改变标准刺激与偏差刺激强度的绝对值,MMN潜伏期没有明显变化;其中只有组4和组6的MMN波幅[(3.41±1.64)μV,(2.37±1.47)μV]差异有统计学意义(P<0.05),其余各组的波幅差异均无统计学意义.结论 MMN只与刺激差异程度有关,与受试者的注意无关,属于认知功能的产物,能够客观地反映大脑探测刺激特征变化的能力.     

Maturation of mismatch negativity in school-age children

OBJECTIVE: Event-related potentials were recorded to investigate the maturation of auditory processing in school-age children. DESIGN: The mismatch negativity (MMN) was obtained in an oddball tone discrimination paradigm in 66 school-age children and 12 adults. In the children's data, a prominent negativity to both the standard and deviant tone, peaking around 200 msec, was observed, and compared with the N1 auditory evoked potential component. RESULTS: The MMN was found to decrease with latency by 11 msec/yr from 4 to 10 yr of age. No developmental change in MMN amplitude was seen from 4 to 10 yr of age. However, the MMN amplitude was significantly smaller in adults than in children. The prominent negativity in children was significantly later than the adult N1 component, and did not change in latency from 4 to 10 yr of age. This finding adds to a body of evidence suggesting that this prominent negativity and the adult N1 are not the same component. The magnitude of the prominent negativity in children decreased slightly with age. CONCLUSION: Changes in the timing of the brain discriminative response, MMN, suggest systematic maturational changes in auditory processing.     

Auditory temporal pattern discrimination and reading ability.

The relation between reading ability and performance on an auditory temporal pattern discrimination task was investigated in children who were either good or delayed readers. The stimuli in the primary task consisted of sequences of tones, alternating between high and low frequencies. The threshold interstimulus interval (ISI) for discrimination of differences in the temporal properties of the sequences was measured. An ISI threshold was also measured in a control task that was identical to the primary task, except all tones in a control sequence had the same frequency. Delayed readers and good readers were equally able to discriminate the timing of the sequences at short ISIs, for both the primary and control tasks. Furthermore, the ISI thresholds were not correlated with the ability to read either irregular words or nonwords. These results suggest that reading ability is not related to the ability to track large and rapid frequency changes in auditory temporal patterns.     

Auditory temporal order discrimination and backward recognition masking in adults with dyslexia.

The ability of 20 adult dyslexic readers to extract frequency information from successive tone pairs was compared with that of IQ-matched controls using temporal order discrimination and auditory backward recognition masking (ABRM) tasks. In both paradigms, the interstimulus interval (ISI) between tones in a pair was either short (20 ms) or long (200 ms). Temporal order discrimination was better for both groups of listeners at long than at short ISIs, but no group differences in performance were observed at either ISI. Performance on the ABRM task was also better at long than at short ISIs and was influenced by variability in masker frequency and by the spectral proximity of target and masker. The only significant group difference was found in one condition of the ABRM task when the target-masker interval was 200 ms, but this difference was not reliable when the measure was of optimal performance. Moderate correlations were observed between auditory thresholds and phonological skill for the sample as a whole and within the dyslexic and control groups. However, although a small subgroup of dyslexic listeners with poor phonology was characterized by elevated thresholds across the auditory tasks, evidence for an association between auditory and phonological processing skills was weakened by the finding of a subgroup of control listeners with poor auditory processing and normal phonological processing skills.     

Event-related potentials in pediatric cochlear implant patients

OBJECTIVE: The main objective of this study was to assess the correlation of auditory event related potential (ERP) measures with behavioral assessment data to identify if ERPs including mismatch negativity (MMN) can be used to categorize cochlear implant patients into good and poor performers. DESIGN: We investigated auditory event-related potentials to standard and deviant speech stimuli presented in a pseudorandom sequence in 35 cochlear implant patients between the ages of 7 and 17 yr. We compared the occurrence, latencies, and amplitudes of P1, N2, and MMN with overall behavioral outcome in these children. Behavioral measures included category of auditory performance scores and speech intelligibility rating scores. RESULTS: Auditory ERPs in response to standard stimuli were identifiable in 30 of 35 patients, demonstrating a major positive component (P1) followed by a negativity (N2) with absence of N1 in all patients. The P1 component in prelingually deaf patients showed a statistically significant reduction in its latency with increasing duration of implant use. MMN was recorded in 80-85% of star performers but in only 15-20% of poor performers. Patients with higher SIR scores demonstrated statistically significant longer duration of MMN compared with those with a lower SIR score. CONCLUSIONS: These results indicate that MMN can be used to assess the functional status of the auditory cortex in terms of auditory memory and discrimination in young children with cochlear implants and may provide an objective mechanism for differentiating good from poor performers.     

Relationship between auditory perception skills and mismatch negativity recorded in free field in cochlear-implant users

This study investigated the ability of cochlear-implanted patients to discriminate tone bursts in free field using the electrophysiological recordings of mismatch negativity (MMN). Seven cochlear-implanted patients (CIP) and eight control subjects (CS) were tested. Event-related potentials were recorded from either 32 or 64 electrodes in response to binaural stimuli using a passive oddball paradigm. Two stimulus-contrast conditions were used to produce MMN: The standard-tone frequency was fixed at 1 kHz, and the deviant-tone frequency was set at 2 or 1.5 kHz. The results show that response waveforms (N1/P2) are similar in latency and amplitude for CS and CIP, suggesting that pure-tone detection is performed over the same time window in both groups. These waveforms are also similar in left- and right-implanted patients, suggesting that electric stimulation of the auditory nerve activates both hemispheres in profound, bilateral hearing loss. Pure-tone audiograms and word-discrimination scores were also measured in each subject in an anechoic room and their relations with MMN data were examined. Correlations were found between the latency of MMN for a 1.5 kHz deviant and the thresholds obtained for pure-tone detection and word discrimination. MMN appears as a possible complementary clinical tool to objectively assess auditory sensitivity in cochlear-implanted populations. However, further improvements are still necessary before it can be used as a standard clinical examination.     

Neural substrate of sound duration discrimination during an auditory sequence in the guinea pig primary auditory cortex

Mismatch negativity (MMN) is a negative component of event-related brain potentials elicited by stimulus transitions. Stimulus duration transition also elicits MMN (duration MMN), with a magnitude that is related to the degree of duration change and the discrimination ability. The neural substrates of duration MMN have not yet been investigated. We therefore studied how duration transitions in an auditory stimulus train are represented in neurons in the primary auditory cortex of anesthetized guinea pigs. Two types of neuronal responses to the context of changes in stimulus duration were found. One was a reduced response as the duration of the preceding stimulus was increased. Second was an enhancement of the late components of the response including sustained and offset responses at the duration transition. The former may be explained by the previously proposed two-tone suppression, which is dependent on the preceding stimulus duration. The latter is likely to be caused by stimulus-specific adaptation that could be a possible neural generator of duration MMN.     

Central auditory nervous system dysfunction in infants with non-syndromic cleft lip and/or palate

Objective

Peripheral hearing loss has been commonly reported in children with non-syndromic cleft lip and/or palate (NSCLP) but few studies have provided information about central auditory nervous system (CANS) functioning for this group. The main objective of this study was to explore CANS functioning in infants with NSCLP through analysis of auditory evoked potentials (AEPs).

Methods

AEPs including auditory brainstem response (ABR), middle latency response (MLR), and mismatch negativity (MMN) recordings were conducted in 34 infants of Chinese ethnicity with NSCLP and an equivalent number of normal controls.

Results

There was no significant difference in ABR (all measurements, including wave I, III, V latencies, I-V inter-wave latency, and wave V amplitude), or MLR (recordable components, Na, Pa latencies, and Na-Pa amplitude) findings between the two groups. However, infants with NSCLP had a significantly smaller MMN response than their normal controls, using MMN strength as the measurement.

Conclusions

Significant abnormal auditory evoked potential findings at the cortical level suggest that infants with NSCLP may be at risk of central auditory discrimination dysfunction. Further effort is needed to determine auditory processing abilities in infants with NSCLP.     

Intracortical mechanisms of mismatch negativity dysfunction in schizophrenia

Event-related potentials provide an objective index of neurocognitive dysfunction in schizophrenia. Schizophrenia subjects show a decreased mismatch negativity (MMN) amplitude relative to age- and sex-matched controls, along with a characteristic pattern of MMN dysfunction across conditions. Deficits in MMN generation are accompanied by (1) impaired precision of auditory sensory memory performance and (2) an interstimulus-interval-dependent deficit in auditory N(1) generation. Similar deficits are observed following systemic or local infusion of N-methyl-D-aspartate (NMDA) antagonists, supporting glutamatergic and phencyclidine/NMDA models of the disorder. Deficits in MMN generation may also be seen following focal cortical damage, especially to the dorsolateral prefrontal cortex. MMN thus provides a useful tool for investigating mechanisms underlying brain dysfunction in schizophrenia.     

Age-related differences in sensitivity to small changes in frequency assessed with cortical evoked potentials

As part of an ongoing study of age-related changes in auditory processing, sensitivity to small changes in frequency were assessed using the cortical auditory evoked potential, P1-N1-P2, in younger and older adults with normal hearing. Behavioral measures have shown age-related differences in intensity and frequency discrimination that are larger at lower than higher frequencies. However, substantial individual differences and equivocal results among studies have been reported. This variability may reflect differences in tasks and procedures, as well as subject variables, such as hearing sensitivity and level of attention. To minimize these subject variables, the P1-N1-P2 response was investigated using a passive listening paradigm. Subjects were 10 younger and 10 older adults. The P1-N1-P2 was elicited by a 150-ms change in frequency in otherwise continuous 500-Hz and 3000-Hz pure tones presented at 70dB SPL. P1-N1-P2 threshold was defined as the smallest change in frequency needed to evoke a P1-N1-P2 response. Furthermore, a frequency-dependent aging effect was observed for P1-N1-P2 thresholds, such that older subjects were significantly less sensitive to the frequency change than younger subjects, with significantly larger age-related differences at 500Hz than at 3000Hz. Age-related changes in response latencies and amplitude of the P1-N1-P2 response were also evident at 500 and 3000Hz. These results are consistent with age-related changes in the central auditory system and suggest that changes in frequency discrimination abilities of older adults may be, in part, related to changes in preattentive levels of auditory processing.     

Acute effect of nicotine on auditory gating in smokers and non-smokers

This paper investigates the role of cholinergic mechanisms in auditory gating by assessing the acute effects of nicotine, an acetylcholinomimetic drug, on behavioral and electrophysiological measures of consonant-vowel (CV) discrimination in quiet and in broadband noise (BBN). In a single-blind procedure, categorical boundaries and mismatch negativity (MMN) in two conditions (quiet, BBN) were obtained from 10 non-smokers and 4 smokers with normal hearing under two drug conditions (nicotine, placebo). After the nicotine sessions, plasma tests revealed a subject's nicotine concentration and subjects reported any symptoms. Larger MMN areas and steeper slopes at the boundary were interpreted as reflecting better electrophysiological and behavioral CV discrimination, respectively. Results indicate that, in non-smokers, the effects of nicotine on electrophysiological CV discrimination in quiet increase with an increase in severity of symptoms. Specifically, asymptomatic non-smokers (N = 5) demonstrate little improvement (and sometimes decrements) in performance while symptomatic non-smokers (N = 5) exhibit nicotine-enhanced discrimination, as do smokers. In noise, all subjects demonstrate nicotine-enhanced behavioral and electrophysiological discrimination. Additionally, in noise, smokers exhibit a larger number of measurable categorical boundaries as well as larger MMN areas than non-smokers in both placebo and nicotine sessions. Results are consistent with the hypothesis that nicotinic cholinergic mechanisms play a role in the gating of auditory stimuli.     

Neural correlates of the precedence effect in auditory evoked potentials

The precedence effect in subjective localization tasks reflects the dominance of directional information of a direct sound (lead) over the information provided by one or several reflections (lags) for short delays. By collecting data of both psychoacoustical measurements and auditory evoked potentials the current study aims at neurophysiological correlates for the precedence effect in humans. In order to investigate whether the stimulus features or the perception of the stimulus is reflected on the ascending stages of the human auditory pathway, auditory brainstem responses (ABRs) as well as cortical auditory evoked potentials (CAEPs) using double click-pairs were recorded. Potentials were related to the results of the psychoacoustical data. ABRs to double click-pairs with lead-lag delays from 0 to 20 ms and interaural time differences (ITDs) in the lag click of 0 and 300 micros show an emerging second wave V for lead-lag delays larger than 2 ms. The amplitudes of the first and second wave V are the same for a lead-lag delay of about 5 ms. For the lag-ITD stimuli the latency of the second wave V was prolonged by approximately ITD/2 compared to the stimuli without lag-ITD. As the amplitudes of the second wave V were not decreased for a lead-lag delay around 5 ms as could be expected from psychoacoustical measurements of the precedence effect, ABRs reflect stimulus features rather than the perceptive qualities of the stimulus. The mismatch negativity (MMN) component of the CAEP for double click-pairs was determined using a diotic standard and a deviant with an ITD of 800 micros in the lag click. The comparison between the MMN components and the psychoacoustical data shows that the MMN is related to the perception of the stimulus, i.e., to the precedence effect. Generally, the findings of the present study suggest that the precedence effect is not a result of a poor sensitivity of the peripheral bottom-up processing. Rather, the precedence effect seems to be reflected by the MMN, i.e., cognitive processes on higher stages of the auditory pathway.     

Statistical detection and analysis of mismatch negativity derived by a multi-deviant design from normal hearing children

To shorten the time of measurement of mismatch negativity (MMN) and to objectify the analysis of MMN characteristics, auditory evoked potentials (AEP) were recorded and MMNs were statistically examined. AEPs from 16 healthy children between the ages of 5 and 7 were elicited using a multi-deviant design including: frequency, duration, intensity, and gap. For detection of MMN an automatic method was applied based on statistical analysis (t-tests) of AEP sweeps. The incidences found were compared to the incidences of MMNs detected by visual inspection of difference waveforms. MMN features such as peak amplitude, peak latency, onset latency, duration of MMN, and area under the curve were evaluated for MMN intervals identified by statistical analysis. MMN incidences identified by statistical analysis (44) reflected a consistency of 84% to visually identified (52) incidences of MMN. Scalp distribution of MMN was mainly fronto-central. Peak latencies of the duration MMN differed significantly from those of the other deviant MMNs. The multi-deviant paradigm was successfully applied for measuring children’s AEPs. The statistical approach proved applicable for analysing the data objectively.