MMN is elicited irrespective of the subject or patient’s attention or behavioural task which implicates the occurrence of an automatic comparison between the current input and the representation, or the memory trace, of the preceding auditory events. This change-detection process occurs unconsciously in the auditory cortices (generating the auditory-cortex subcomponent of the MMN). However, it activates, with a very short time delay, frontal-cortex mechanisms (generating the frontal subcomponent of the MMN) controlling the direction of attention, which leads to attention switch to, and conscious perception of, sound change . Thus, the MMN is also involved in initiating a cerebral warning mechanism which is of great biological significance.
While the basic mechanism of the MMN is simple, with appropriate experimental manipulations, it is a versatile tool for investigating various aspects of auditory perception and attention. Here are some examples on the applicability of the MMN for neuroscience studies:
Traditionally, the MMN has been recorded with an oddball paradigm, which typically includes a repetitive standard stimulus (e.g., a 1000 Hz tone, p. 0.9) occasionally replaced by a deviant stimulus (e.g., 1100 Hz, p. 0.1). This approach is very time consuming, always having the trade-off of MMN signal quality and the amount of information obtained (the number of different deviant types for which the MMN is recorded). Especially for investigating patient groups and young children, this approach is not optimal since the recording times should be kept minimal. At the same time, the EEG trial loss may be high due to, e.g., movements.
To overcome these problems, Näätänen et al.,  developed a new multi-feature MMN paradigm, called initially “Optimum-1”, with which MMN for five different types of deviant sounds can be recorded in the same time as for one deviant type in the oddball paradigm described above. The deviants included in such multi-feature sequence are alternating with the standard stimulus, and each deviant stimulus should differ from the rest of the stimuli in one feature only. The rationale of this paradigm is that besides serving as a deviant, each deviant stimulus type also strengthens the memory trace for the features it shares with the rest of the stimuli, thereby acting as a “standard”. For example, if only the frequency of a sound changes, this sound still strengthens the memory traces for sound duration, intensity, and location.