A number of instruments can be used to help characterize the behavior of the glottis and vocal folds during phonation. The signals derived from these instruments are called glottographic waveforms or glottograms (Titze and Talkin, 1981). Among the more common glottograms are those that track change in glottal flow, via inverse filtering; glottal width, via kymography; glottal area, via photoglottography; and vocal fold movement, via ultrasonography (Baken and Orlikoff, 2000). Such signals can be used to obtain several different physiological measures, including the glottal open quotient and the maximum flow declination rate, both of which are highly valuable in the assessment of vocal function. Unfortunately, the routine application of these techniques has been hampered by the cumbersome and time-consuming way in which these signals must be acquired, conditioned, and analyzed. One glottographic method, electroglottography (EGG), has emerged as the most commonly used technique, for several reasons: (1) it is noninvasive, requiring no probe placement within the vocal tract; (2) it is easy to acquire, alone or in conjunction with other speech signals; and (3) it offers unique information about the mucoundulatory behavior of the vocal folds, which contemporary theory suggests is a critical element in the assessment of voice production.

Electroglottography (known as electrolaryngography in the United Kingdom) is a plethysmographic technique that entails fixing a pair of surface electrodes to each side of the neck at the thyroid lamina, approximating the level of the vocal folds. An imperceptible low-amplitude, high-frequency current is then passed between these electrodes. Because of their electrolyte content, tissue and body fluids are relatively good conductors of electricity, whereas air is a particularly poor conductor. When the vocal folds separate, the current path is forced to circumvent the glottal air space, decreasing effective voltage. Contact between the vocal folds affords a conduit through which current can take a more direct route across the neck. Electrical impedance is thus highest when the current path must completely bypass an open glottis and progressively decreases as greater contact between the vocal folds is achieved. In this way, the voltage across the neck is modulated by the contact of the vocal folds, forming the basis of the EGG signal. The glottal region, however, is quite small compared with the total region through which the current is flowing. In fact, most of the changes in transcervical impedance are due to strap muscle activity, laryngeal height variation induced by respiration and articulation, and pulsatile blood volume changes. Because increasing and decreasing vocal fold contact has a relatively small effect on the overall impedance, the electroglottogram is both high-pass filtered to remove the far slower nonphonatory impedance changes and amplified to boost the laryngeal contribution to the signal. The result is a waveform—sometimes designated Lx—that varies chiefly as a function of vocal fold contact area (Gilbert, Potter, and Hoodin, 1984).

First proposed by Fabre in 1957 as a means to assess laryngeal physiology, the clinical potential of EGG was recognized by the mid-1960s. Interest in EGG increased in the 1970s as the importance of mucosal wave dynamics for vocal fold vibration was confirmed, and accelerated greatly in the 1980s with the advent of personal computers and commercially available EGGs that were technologically superior to previous instruments. Today, EGG has a worldwide reputation as a useful tool to supplement the evaluation and treatment of vocal pathology. The clinical challenge, however, is that a valid and reliable EGG assessment demands a firm understanding of normal vocal fold vibratory behavior along with recognition of the specific capabilities and limitations of the technique.

Instead of a simple mediolateral oscillation, the vocal folds engage in a quite complex undulatory movement during phonation, such that their inferior margins approximate before the more superior margins make contact. Because EGG tracks effective medial contact area, the pattern of vocal fold vibration can be characterized quite well (Fig. 1). The contact pattern will vary as a consequence of several factors, including bilateral vocal fold mass and tension, medial compression, and the anatomy and orientation of the medial surfaces. Considerable research has been devoted to establishing the important features of the EGG and how they relate to specific aspects of vocal fold status and behavior. Despite these efforts, however, the contact area function is far from perfectly understood, especially in the face of pathology. Given the complexity of the “rolling and peeling” motion of the glottal margins and the myriad possibilities for abnormality of tissue structure or biomechanics, it is not surprising that efforts to formulate simple rules relating abnormal details to specific pathologies have not met with notable success. In short, the clinical value of EGG rests in documenting the vibratory consequence of pathology rather than in diagnosing the pathology itself.

Figure 1..  

At the top is shown a schematic representation of a single cycle of vocal fold vibration viewed coronally (left) and superiorly (right) (after Hirano, 1981). Below it is a normal electroglottogram depicting relative vocal fold contact area. The numbered points on the trace correspond approximately to the points of the cycle depicted above. The contact phases of the vibratory cycle are shown beneath the electroglottogram.

Using multiple glottographic techniques, Baer, Löfqvist, and McGarr (1983) demonstrated that, for normal modal-register phonation, the “depth of closure” was very shallow just before glottal opening and quite deep soon after closure was initiated. Most important, they showed that the instant at which the glottis first appears occurs sometime before all contact is lost, and that the instant of glottal closure occurs sometime after the vocal folds first make contact. Thus, although the EGG is sensitive to the depth of contact, it cannot be used to determine the width, area, or shape of the glottis. For this reason, EGG is not a valid technique for the measurement of glottal open time or, therefore, the open quotient. Likewise, since EGG does not specify which parts of the vocal folds are in contact, it cannot be used to measure glottal closed time, nor can it, without additional evidence, be used to determine whether maximal vocal fold contact indeed represents complete obliteration of the glottal space. Identifying the exact moment when (and if) all medial contact is lost has also proved particularly problematic. Once the vocal folds do lose contact, however, it can no longer be assumed that the EGG signal conveys any information whatsoever about laryngeal behavior. During such intervals, the signal may vary solely as a function of the instrument's automatic gain control and filtering (Rothenberg, 1981).

Although the EGG provides useful information only about those parts of the vibratory cycle during which there is some vocal fold contact, these characteristics may provide important clinical insight, especially when paired with videostroboscopy and other data traces. EGG, with its ability to demonstrate contact change in both the horizontal and vertical planes, can quite effectively document the normal voice registers (Fig. 2) as well as abnormal and unstable modes of vibration (Fig. 3). However, to qualitatively assess EGG wave characteristics and to derive useful indices of vocal fold contact behavior, it may be best to view the EGG in terms of a vibratory cycle composed of a contact phase and a minimal-contact phase (see Fig. 1). The contact phase includes intervals of increasing and decreasing contact, whereas the peak represents maximal vocal fold contact and, presumably, maximal glottal closure. The minimal-contact phase is that portion of the EGG wave during which the vocal folds are probably not in contact. Much clinical misinterpretation can be avoided if no attempt is made to equate the vibratory contact phase with the glottal closed phase or the minimal-contact phase with the glottal open phase.

Figure 2..  

Typical electroglottograms obtained from a normal man prolonging phonation in the low-frequency pulse, moderate-frequency modal, and high-frequency falsetto voice registers.

Figure 3..  

Electroglottograms representing different abnormal modes of vocal fold vibration.

For the typical modal-register EGG, the contact phase is asymmetrical; that is, the increase in contact takes less time than the interval of decreasing contact. The degree of contact asymmetry is thought to vary not only as a consequence of vocal fold tension but also as a function of vertical mucosal convergence and dynamics (i.e., phasing; Titze, 1990). A dimensionless ratio, the contact index (CI), can be used to assess contact symmetry (Orlikoff, 1991). Defined as the difference between the increasing and decreasing contact durations divided by the duration of the contact phase, CI will vary between −1 for a contact phase maximally skewed to the left and +1 for a contact phase maximally skewed to the right. For normal modal-register phonation, CI varies between −0.6 and −0.4 for both men and women, but, as can be seen in Figure 2, it is markedly different for other voice registers. Pulse-register EGGs typically have CIs in the vicinity of −0.8, whereas in falsetto it would not be uncommon to have a CI that approximates zero, indicating a symmetrical or nearly symmetrical contact phase.

Another EGG measure that is gaining some currency in the clinical literature is the contact quotient (CQ). Defined as the duration of the contact phase relative to the period of the entire vibratory cycle, there is evidence from both in vivo testing and mathematical modeling to suggest that CQ varies with the degree of medial compression of the vocal folds (see Fig. 3) along a hypoadducted “loose” (or “breathy”) to a hyperadducted “tight” (or “pressed”) phonatory continuum (Rothenberg and Mahshie, 1988; Titze, 1990). Under typical vocal circumstances, CQ is within the range of 40%–60%, and despite the propensity for a posterior glottal chink in women, there does not seem to be a significant sex effect. This is probably due to the fact that EGG (and thus the CQ) is insensitive to glottal gaps that are not time varying. Unlike men, however, women tend to show an increase in CQ with vocal F0. It has been conjectured that this may be the result of greater medial compression employed by women at higher F0s that serves to diminish the posterior glottal gap. Nonetheless, a strong relationship between CQ and vocal intensity has been documented in both men and women, consistent with the known relationship between vocal power and the adductory presetting of the vocal folds. Because vocal intensity is also related to the rate of vocal fold contact (Kakita, 1988), there have been some preliminary attempts to derive useful EGG measures of the contact rise time.

Because EGG is relatively unaffected by vocal tract resonance and turbulence noise (Orlikoff, 1995), it allows evaluation of vocal fold behavior under conditions not well-suited to other voice assessment techniques. For this reason, and because the EGG waveshape is a relatively simple one, the EGG has found some success both as a trigger signal for laryngeal videostroboscopy and as a means to define and describe phonatory onset, offset, intonation, voicing, and fluency characteristics. In fact, EGG has, for many, become the preferred means by which to measure vocal fundamental frequency and jitter.

In summary, EGG provides an innocuous, straightforward, and convenient way to assess vocal fold vibration through its ability to track the relative area of contact. Although it does not supply valid information about the opening and closing of the glottis, the technique affords a unique perspective on vocal fold behavior. When conservatively interpreted, and when combined with other tools of laryngeal evaluation, EGG can substantially further the clinician's understanding of the malfunctioning larynx and play an effective role in therapeutics as well.

See also acoustic assessment of voice.