Ms. Brown is currently enrolled in her second year of the audiology doctorate program at the University of Connecticut in Storrs. Dr. Musiek, the clinical editor of Pathways, is a professor and the director of auditory research in communication sciences at the University of Connecticut.
The masking level difference (MLD) is typically a threshold measure, and it allows easier detection of tone or speech in the presence of noise as a result of changing the phase of the signal or the masker. It is usually expressed in decibels as the difference between a homophasic, or the in-phase condition, and an antiphasic condition, in which the signal or the noise are out of phase.
This psychoacoustic phenomenon was described for speech and tones in 1948. (J Acoust Soc Am 1948;20:150; J Acoust Soc Am 1948;20:536.) Numerous studies have since revealed that phase manipulation can have a significant effect on audibility. A standard masked threshold can be determined when one ear is presented with noise and a signal. The signal becomes easier to hear if that noise is presented binaurally. The listening situation once again becomes difficult if the signal and noise are presented to both ears.
The listening situation becomes much easier, however, when the signal is put 180 degrees out of phase between the ears, while the noise remains in phase, or when the noise is 180 degrees out of phase between ears while the signal remains in phase. This improvement in detection is the result of complex binaural interaction and can provide valuable information regarding the auditory system. It is frequency dependent, with the greatest release from masking occurring in the low frequencies.
Test conditions have different notations. “S” represents the signal, “N” represents the noise or masker, “m” refers to being presented to one ear, or monaurally, “o” refers to being in phase at the two ears and “π” refers to being 180 degrees out of phase between the two ears. The noise typically used as the masker is either broad-spectrum noise or narrowband noise centered on the test frequency. Different conditions and nomenclature have been utilized in determining the MLD. (See table.)
MLD into Practice
Masking level differences are determined by calculating the difference in threshold between an antiphasic condition and a homophasic condition. A lower threshold for the antiphasic condition demonstrates an increase in detectability and therefore a release from masking. The binaural MLD procedure for a 500-Hz tone would consist of determining the threshold in the presence of an effective masking level of noise when the signal and the noise are presented in phase between the ears.
Assume, for example, this condition yields a threshold of 61 dB HL. The phase of either the masker or the signal would then be changed 180 degrees out of phase between the ears. This example demonstrates a 12-dB release from masking, or masking level difference, if this phase change results in a threshold of 49 dB HL.
The masking level difference for normal listeners ranges from 8 to 12 dB, varying slightly depending on the type of signal being used and whether the noise or signal is being manipulated out of phase. (Arch Otolaryngol 1981;107:357.) The antiphasic condition in which the signal is out of phase between ears and the noise is in phase appears to result in the largest MLD. Typically, studies consider an MLD below 7 dB as abnormal. (Ann Otol Rhinol Laryngol 1976;85[6 Pt 1]:820; Arch Otolaryngol 1977;103;482.) Studies have shown a decrease in MLD as the result of peripheral hearing loss. Sensorineural and conductive hearing losses have been shown to demonstrate a reduced MLD. Cochlear hearing loss may have an effect on the MLD because of the system's inability to maintain temporal cues. Overall, an asymmetry in hearing sensitivity seems to result in a reduced ability to incorporate these binaural cues.
Researchers have demonstrated that this phenomenon appears to be mediated in the central auditory nervous system below the level of the auditory cortex. This is supported by patients with cortical pathology who consistently demonstrate MLDs comparable with normal listeners. MLDs appear to be most sensitive and are most reduced when brainstem lesions or disease is present.
Multiple sclerosis is a central pathology that affects the brainstem and has been extensively studied using the MLD. Numerous studies have shown abnormal MLD levels in this population. The eighth cranial nerve tumors and lesions have also been shown to result in reduced MLDs. The region of the superior olivary complex in the caudal pons appears to be a likely site to carry out this release from masking function because it is the first location in the auditory system that has contralateral representation. (Arch Otolaryngol 1981;107:357.)
Studies show that MLD is a good behavioral test for assessing auditory function of the lower brainstem. The fact that it is better at detecting brainstem disease, in particular in the lower brainstem, makes MLD a good indicator of central auditory involvement. It can be a valuable part of the central auditory system diagnostic test battery.