Hearing Research.

29 (1987) 33-43 33
Elsevier

HRR 00940

The human auditory brain stem as a generator of auditory evoked potentials

Jean K. Moore
Dept.ofAnatomical Sciences, SUNY at Stony Brook, Stony Brook, New York, U.S.A.

(Received 27 May 1986; accepted 12 March 1987)

Data on the size, location, and orientation of human brain stem auditory nuclei are discussed here from the point of view of the
potential role of these structures in generation of the brain stem auditory evoked response. Due to reduction in size of several nuclei
in the human brain stem, the structures most likely to be generators of far-field potentials are the cochlear nuclei, medial olivary
nuclei, dorsal lemniscal nuclei, inferior colliculi, and their axonal projections. Consideration of the anatomy of the human auditory
brain stem pathway forms the basis for a model of generation of auditory evoked potentials.

Human brain stem; Auditory evoked potential

Introduction (Wada and Starr, 1983a,b,c). A basic premise un-

derlying use of these studies has been that human
Auditory brain stem evoked responses (ABRs) auditory centers are essentially identical to those
are a series of low-amp~tude positive and negative of carnivores and rodents. However, several prior
potentials recorded from the scalp within 10 ms anatomical studies of human brain stem nuclei
following a brief click or tone stimulus. The hu- (Moore and Moore, 1971; Strominger and
man ABR is used clinically as a measure of brain Hurwitz, 1976; Moore and Osen, 1979) suggest
stem integrity and matu~ty (Hecox and Galam- that this assumption of total similarity of human
bos, 1974) and also as a diagnostic tool in identifi- and other mammalian auditory centers is not en-
cation of specific brain stem pathology (Starr and tirely valid. Further, a recent comparative study of
Hamilton, 1976; Stockard and Rossiter, 1977; the cat and human brain stem pathways (Moore,
Hashimoto et al., 1979). 1987a) reveals that the human lower auditory sys-
The component waves of the ABR are assumed tem is in some respects very similar to that of
to have been generated by the cochlear nerve and other mammals, but in other respects strikingly
subsequent structures of the ascending brain stem different.
auditory pathway. Interpretation of the ABR has This article considers the topography and cyto-
been based to a large extent on hypotheses of architecture of human auditory centers from the
wave generation developed in experimental animal standpoint of their role as generators of the ABR.
studies. These studies include depth electrode re- Information on size, location, and other morpho-
cordings and lesion studies done in the brain logical characteristics of these centers is correlated
stems of the cat (Buchwald and Huang, 1975; with results of recordings of evoked potentials
Huang and Buchwald, 1977; Achor and Starr, from the surface of the eighth nerve and brain
1980a,b), mouse (Henry, 1979), and guinea pig stem in neurosurgical patients (~ashimoto et al.,
1981; Merller et al., 1981; Moller and Jannetta,
1981. 1982a,b,c). These correlations of human
Correspondence ro: Jean K. Moore, Department of Anatomical anatomy with the surface location of evoked
Sciences, SUNY at Stony Brook, Stony Brook, NY 11794, potentials form the basis for a tentative model of
U.S.A. generation of the ABR in man.

037%5955/87/$03.50 0 1987 Elsevier Science Publishers B.V. (Biomedical Division)

34

Topography of the human brainstem auditory path-

way

A somewhat schematic representation of the

human brain stem auditory pathway, as seen in a
horizontal, or overhead, view, is shown in Fig. 1.
This figure is based on the reconstruction of the
brainstem of a 62 year old male from closely
spaced cell- and fiber-stained frontal sections
(Moore, 1987a). It illustrates the fact that the
human lower auditory pathway is composed of the
same major subdivisions as in other mammals, in
that it consists of a co&ear nuclear complex,
superior olivary complex, lemniscal nuclei, and
inferior colliculus. At the same time, this figure
depicts some of the uniquely human topographic
features of the central auditory centers. The
co&ear nuclei (CN) lie in the same position as in
other species, at the lateral edge of the pontome-
dullary junction, but enlargement of the underly-
ing inferior and middle cerebellar peduncles has Fig. 1. In this schematic representation of a horizontal view of
the human brain stem auditory pathway, heavy lines mark the
caused a rather extreme mediolateral elongation of
brain stem surface, with a dashed line indicating the cut
the nuclei, with outward rotation of the rostra1 surface of the middle cerebellar peduncle (MCP). Thin lines
end of the complex. The main body of the super- indicate nuclear groups, and dotted lines show the location of
ior olivary complex (SOC) lies deeper in the pons, axonal pathways. The cochlear nuclei (CN) are a mediolater-
a relatively short distance medial and rostra1 to ally elongated cell group lying at the lateral edge of the brain
the cochlear nuclei. Axons from the cochlear nuclei at the level of the pontomedullary junction. The most likely
course of the efferent axons from the cochlear nuclei in the
reach both the ipsilateral and contralateral olivary trapezoid body (TB) is indicated. The main body of the super-
complexes through an axonal pathway, the ior olivary complex (SW) is located in the pons a short
trapezoid body (TB). The much smaller dorsal distance rostra1 and medial to the co&ear complex: a small
efferent pathway, the acoustic stria, is present in number of commissural axons may arise from cells in the
human brain stem but is not illustrated in this olivary nuclei. From the main body of the complex, a column
of periolivary cells extends forward in the ascending axonal
diagram. The superior olivary complex in most pathway. At the rostral edge of the middle cerebellar peduncle,
mammals is a compact group of nuclei which the pathway moves laterally to the surface of the pons to form
abuts directly rostrally on the lemniscal nuclei. In the classical lateral lemniscus. This segment of the pathway
the human brain stem, however, the much longer contains the nuclei of the lateral lemniscus (NLL). The only
pontine segment of the brainstem creates an traces of a ventral lemniscal nucleus are scattered small areas
of neuropil, but the dorsal lemniscal nucleus is a distinct cell
elongated rostra1 pole of the olivary complex which group which is the origin of the dorsal commissure of the
extends forward in the auditory pathway through lemniscus (DCLL). The lemniscus terminates in the inferior
most of the posterior half of the pons. At the colliculus (IQ paired midbrain structures which are intercon-
point where the pons is no longer covered laterally nected the commissure of the inferior colliculus (CIC).
by the middle cerebellar peduncle (MCP), the
auditory pathway swings laterally to become the nucleus is a distinct cell group which gives rise to
flattened, superficial band of axons traditionally the dorsal commissure of the lateral lemniscus
referred to as the lateral lemniscus. Within this (DCLL). The human lateral lemniscus, like that of
segment of the pathway lie the cell groups con- other species, terminates in the inferior colliculus
stituting the nuclei of the lateral lemniscus (NLL). (IC) of the midbrain &turn. Rostrally, the two
The human ventral lemniscal nucleus is reduced to inferior colliculi are interconnected through the
scattered small clusters of cells, but the dorsal collicular commissure (CIC). Many of the topo-
 35

graphical features illustrated here have been de- Unpubl.). Thus a response in the human brain
scribed by Dublin (1976) and are depicted in a stem to a transient stimulus will consist of waves
reconstruction of the human auditory pathway of action potentials passing along a group of
produced by Kiang and coworkers (1984). relatively thin myelinated axons for a total dis-
A schematic representation such as that in Fig. tance of at least 34 to 4f cm. In addition to the
1 can give the impression that the brain stem time of axonal conduction, the most direct brain
auditory pathway consists of a series of nuclei stem pathway will involve two synaptic delays,
activated sequentially, i.e., that auditory informa- one in the cochlear nuclei and one in the inferior
tion is processed by chains of neurons relaying in colliculus, while relayed pathways will add an
turn in the cochlear nuclei, olivary complex, additional one or two synaptic delays in the olivary
lemniscal nuclei, and finally inferior colliculus. A and lemniscal nuclei. These facts strongly suggest
large body of experimental investigation in animals a total time frame of several milliseconds for
indicates that this is not the case. The only obliga- transmission through the central brain stem
tory synaptic stations in the brainstem pathway pathway. Stated in terms of relative latency, the
are the cochlear nuclei and the inferior colliculus. earliest evoked potentials in the inferior colliculus
Between these two nodal points run a series of should follow the evoked response of the cochlear
pathways in parallel. Some of these pathways ex- nuclei by at least several milliseconds.
tend from the cochlear nuclei directly to the infe-
rior colliculus. Others are relayed through one, or Components of the human auditory pathway
less often two, of the intervening olivary and
lemniscal nuclei. One central issue, to be consid- A major conclusion from the comparison of the
ered in a subsequent section, is the question of cat and human brain stem pathways (Moore,
which of the multiple auditory centers and 1987a) is that there are a number of instances of
pathways remain prominent in the human brain. reduction in size or loss of nuclei and cell groups
All channels between the cochlear nuclei and in the human brain. From the standpoint of audi-
the inferior colliculus, whether direct or relayed tory centers as generators of the ABR, any reduc-
through pontine auditory nuclei, must traverse a tion in the number of auditory nuclei and tracts in
certain minimum distance through the brain stem. the human brain stem tends to simplify the task of
In animal models, conduction time of axonal correlation of human anatomy with the ABR
potentials has not generally been considered to be waveform. In making such a correlation, it is
a significant factor in elapsed time of auditory necessary to consider which auditory centers are
evoked potentials. Since the distance between the still present in the human brain stem and of
cochlear nuclei and the inferior colliculus in the sufficient size to produce activity detectable as a
cat is 14 mm in the ipsilateral pathway and 22 mm far-field response. It is also necessary to consider
contralaterally (Moore, 1987a), and the same how such morphological features as dendritic
pathways in rodent brains must be even shorter, orientation are likely to enhance evoked re-
axonal conduction time in these species may well sponses. In the following discussion, it is taken as
be a relatively unimportant factor. However, the axiomatic - an unprovable assumption - that nuclei
much longer human brain stem creates a signifi- in the human brain stem retain the same afferent
cantly longer ascending auditory pathway in man. and efferent connections as those demonstrated in
Measurements made on the reconstruction of a other species.
human brain stem (Moore, 1987a) indicate that
cochlear axons traverse a distance of about 35 mm Cochlear nuclei
to reach the center of the ipsilateral inferior col- The human cochlear complex is elongated. ex-
liculus and at least 46 mm to the center of the tending about 8 mm mediolaterally but only 3 mm
contralateral colliculus. The human lateral or less rostrocaudally (Terr and Edgerton, 1985;
lemniscus in myelin-stained sections is seen to be Moore, 1987a). A quantitative study of the com-
a relatively homogeneous population of lightly plex (Hall, 1965) has determined that it contains
myelinated axons, 2-4 pm in diameter (Moore. close to 100000 neurons. Assuming that a major-
ity of these neurons are innervated by the co&ear
nerve, an auditory stimulus could evoke depolari-
zation of most of the neurons in the complex. One
factor which has been considered important in
determining the effectiveness of neural centers as
dipole generators is some sort of parallel orienta-
tion of their dendrites. In the human cochlear
nuclei, Nissl and Bodian studies (Dublin, 1976;
Moore, 1987a) show a tendency toward parallel
orientation of dendrites in the dorsal nucleus, but
dendritic orientation in the ventral nucleus is more
B.
or less random. Nonetheless, even in the absence
of an overall consistent parallelism of dendrites,
the firing of this large population of cochlear Fig. 2. This figure presents a wiring diagram of the major
neurons should be an event clearly detectable in auditory pathways in the human brain stem. (A) The cochlear
far-field recordings. nuclei (CN) project to the inferior colliculus (IC), with a much
heavier projection to the contralateral side of the brain. En
A direct projection from the co&ear nuclei to route to the colliculus, this pathway provides some input to the
the inferior colliculus has been demonstrated in dorsal nucleus of the lemniscus on the contralateral side. The
the cat (Warr, 1966, 1969; Fernandez and Kara- dorsal lemniscal nucleus has an ipsilateral projection to the
pas, 1967; Osen, 1972; Adams, 1979; Brunso-Be- inferior colliculus and a commissural projection to the opposite
dorsal nucleus and colliculus. (B) The cochlear nuclei project
chtold et al., 1981), rat (Beyerl, 1978), mouse
to both medial olivary nuclei (MSO). Under conditions of
(Ryugo et al., 1981), gerbil (Nordeen et al., 1983), monaural stimulation, this pathway is more active ipsilaterally.
guinea pig (Moore, 1987b), bat (Zook and Casse- Each medial olivary nucleus projects to the dorsal lemniscal
day, 1980; Schweizer, 1981), monkey (Strominger nucleus and to the inferior colliculus on the same side of the
and Strominger, 1971; Strominger, 1973), and brain.
chimpanzee (Strominger et al., 1977). This projec-
tion has consistently been described as heavily or
entirely contralateral, and quantitative studies in can be identified (Moore and Moore, 1971;
the cat and gerbil (Brunso-Bechtold et al., 1981; Strominger and Hurwitz, 1976; Dublin, 1976;
Nordeen et al., 1983) confirm that more than 90% Moore, 1987a). The human medial olivary nucleus,
of the direct cochleocollicular axons project to the in contrast, is a very prominent laminar nucleus,
opposite colliculus. Thus it seems likely that this containing 10 000-12 000 neurons and extending 4
projection is also a predominantly crossed system mm rostrocaudally through the main body of the
in the human brain stem. One consequence of olivary complex (Moore and Moore, 1971;
activation of neurons in the cochlear nuclei by a Strominger and Hurwitz, 1976; Dublin, 1976;
transient stimulus could be a wave of action Moore, 1987a). Periolivary cells are clearly present
potentials traveling in a sizeable population of in the human olivary complex, ringing the medial
axons for roughly 44 cm to reach the central nucleus and forming a column of cells extending
nucleus of the inferior colliculus on the opposite roughly 8 mm rostrally (Fig. 1). Though there are
side of the brain stem (Fig. 2A). probably several thousand periolivary cells in the
human brain stem, these cells are scattered over a
Superior olivary complex segment of the pathway more than a centimeter in
In the cat, the superior olivary complex consists length, and it seems unlikely that such a dispersed
of three roughly equal sized nuclei - the medial group of cells would give rise to a sharply de-
olivary nucleus, lateral olivary nucleus, and nucleus marcated potential. In summary, the major event
of the trapezoid body - surrounded by a ring of in activation of the human superior olivary com-
periolivary cell groups. In the human olivary com- plex should be the firing of its medial nucleus. In
plex, the lateral olivary nucleus is a small, indis- addition to being the largest component of the
tinct cell group and no distinct trapezoid nucleus human olivary complex, the medial nucleus has a
distinctive morphology which should make it an but with monaural stimulation, the ipsilateral me-
effective dipole generator. Golgi studies of this dial nucleus is more active (Masterton and Imig,
nucleus (Ramon y Cajal, 1909; Scheibel and 1984). Thus, under the standard monaural condi-
Scheibel, 1974; Schwartz, 1977; 0110 and Schwartz, tions of recording auditory evoked potentials, the
1979) have illustrated that a striking feature of the pathway relayed through the ipsilateral medial
medial nucleus is its long straight dendrites ex- nucleus should be more active than the con-
tending medially and laterally from the central tralateral pathway (Fig. 2B).
column of cell bodies. The pronounced parallel
orientation of the dendrites should enhance the Nuclei of the lateral lemniscus
efficacy of the medial nucleus as a generator, and The human ventral lemniscal nucleus is re-
in fact, all depth electrode studies of the olivary duced to a few clusters of small cells in the
complex in animals have noted a large field poten- patches of neuropil scattered throughout the lateral
tial upon stimulation of the afferent axons to the lemniscus as it passes through the upper pons. In
olivary complex. This potential has been shown to contrast, the dorsal lemniscal nucleus is a promi-
surround the medial olivary nucleus and is attri- nent cell group, very similar in size and mor-
buted to depolarization of its parallel dendritic phology to the corresponding nucleus in the cat
arbors (Galambos et al., 1959; Tsuchitani and (Dublin, 1976; Moore, 1987a). No cell counts are
Boudreau, 1964; Goldberg and Brown, 1968; Clark available for the dorsal lemniscal nucleus. but as a
and Dunlop, 1968; Guinan et al., 1972). This field nucleus 1 mm to 2 mm in diameter, with an
potential should considerably enhance the re- irregularly clustered pattern of distribution of its
sponse of the medial nucleus in the far-field re- neurons, it is a considerably smaller structure than
cording. the medial olivary nucleus. Many, but not all, of
The two medial olivary nuclei are symmetri- its neurons and their dendritic arbors are oriented
cally innervated by axons from the cochlear nuclei horizontally, parallel to the axons of the dorsal
(Harrison and Irving, 1966; Warr, 1966, 1982; commissure (Kane and Barone, 1980). Despite
Strominger and Strominger, 1971) with collaterals this potential for dipole generation, however, the
of the same axons reaching both nuclei small size of the dorsal lemniscal nucleus suggests
(Schneiderman and Henkel, 1985). Action poten- that its firing would be a relatively minor event in
tials passing along these axons travel 9 mm to ABR recording.
reach the ipsilateral medial nucleus and 18 mm to In the cat, afferent input to the dorsal nucleus
reach the contralateral nucleus (Moore, 1987a). comes from all lower brain stem auditory nuclei in
This bilateral innervation means that there should a pattern that mirrors that to the inferior collicu-
be a nearly simultaneous response of both medial lus (Glendenning et al., 1981). In the human brain,
olivary nuclei, a total of 20000-24000 neurons, its major sources of input should thus be the
firing within a short time after the response in the contralateral cochlear nuclei and both medial
cochlear nuclei. Each medial olivary nucleus pro- olivary nuclei (Figs. 2A, B). The fact that afferent
jects virtually exclusively to the inferior colliculus impulses can reach the dorsal nucleus over several
on the same side of the brain (Beyerl, 1978; Adams, routes should tend to make its evoked response
1979; Brunso-Bechtold et al., 1981; Nordeen et al., somewhat broader and less well defined. The pro-
1983; Henkle and Spangler, 1983). Thus depolari- jections from the nucleus are also complex, as its
zation of the medial nuclei should set off a wave axons terminate in the inferior colliculus on the
of action potentials in each lateral lemniscus. The same side, and, through the dorsal commissure, in
symmetrical connections of the medial nucleus the contralateral dorsal lemniscal nucleus and in-
leads to the expectation that this pathway would ferior colliculus (Goldberg and Moore. 1967;
be equally active bilaterally, but 2-deoxyglucose Kudo, 1981).
studies indicate that this pathway is not neces-
sarily functionally symmetrical. Under conditions Inferior colliculus
of binaural stimulation, the medial nucleus is most The inferior colliculus is the largest single com-
active on the side opposite a lateralized stimulus, ponent of the human brain stem auditory pathway.
It is an irregularly spherical nucleus, roughly 6-7 Correlation of brainstem anatomy and ABR wave-
mm in diameter in all planes of section. Cell form
density is high in the central nucleus and
dorsomedial area, and slightly less in the peri- In man, a brain stem auditory system consist-
central region. Although there have been no cell ing mainly of a co&ear nuclear complex, medial
counts of the colliculus, the total number of neu- olivary nucleus, dorsal lemniscal nucleus, and in-
rons in this structure should far outnumber that in ferior colliculus is the neural generator of the
any of the lower auditory centers. One well known ABR. However, even quantitative information on
feature of the morphology of the central nucleus the structures which are the neural sources of
of the colliculus is its fibrodendritic laminae early evoked auditory potentials does not in itself
formed by alternating sheets of axons and neurons provide a complete answer to the issue of genera-
with flattened disc-shaped dendritic fields (Morest tion of the ABR. One remaining question is that
and Oliver, 1984). Similar layers and classes of of identifying the initial brain stem response, i.e.,
neurons are present in the central nucleus of the determining which part of the ABR is generated
human colliculus (Geniec and Morest, 1971). At by the eighth nerve and which part by central
the same time, there are several classes of neuron brain stem structures. A second question is how
in the central nucleus with more symmetrical den- activity in human central auditory structures is
dritic arbors, and neurons in the pericentral areas reflected in the positive and negative waves which
are generally multipolar with radiating dendrites. make up the ABR. Although there is general
In sum, inferior colliculus cytoarchitecture in- agreement that the synchronized activity of popu-
cludes both laminar and nonlaminar features, but lations of neurons and their axons produce the
the parallel orientation of the disc-shaped den- scalp-recorded evoked response, no consensus has
dritic fields in the central nucleus should enhance been reached as to how graded somatodendritic
the potential generated by the colliculus. potentials and axonal action potentials are re-
The inferior colliculus is an obligatory synapse flected in the positive and negative deflections
for the overwhelming majority (> 99%) of the constituting the ABR.
axons ascending in the lateral lemniscus (Gold- The present study of the human brain stem
berg and Moore, 1967; Aitkin and Phillips, 1984). (Moore, 1987a) does not contribute any informa-
This means that it is a target for all of the individ- tion toward the issue of the segment of the ABR
ual pathways from lower auditory nuclei, as il- which is generated by the cochlear nerve. It is
lustrated in Fig. 2. The shortest route to the necessary to consider the work of other investiga-
colliculus would be that from the ipsilateral tors in order to arrive at some reasonable estimate
cochlear nuclei, about 35 mm in length. Fiber of the time of the first evoked activity in the
traveling from the contralateral cochlear nuclei cochlear nuclei, since this response is the baseline
would travel about 46 mm. Both ipsilaterally and against which the latencies of other central re-
contralaterally, axons passing through the dorsal sponses must be measured. In both human and
commissure would carry impulses for approxi- animals studies, there is good agreement that wave
mately an additional 15 mm to reach the collicu- I is identical to the Nl potential recorded from
lus. This convergence of pathways of different the round window of the cochlea and that it
lengths and involving different numbers of syn- reflects intracochlear neural activity (Sohmer and
aptic junctions should activate a large number of Feinmesser, 1967; Lev and Sohrner, 1972; Picton
neurons in the colliculus, but with considerably et al., 1974; Starr and Achor, 1975; Buchwald and
less synchrony in their firing than occurs in either Huang, 1975; Stockard and Rossiter, 1977; Henry,
the cochlear nuclei or the medial olivary nucleus. 1979; Achor and Starr, 1980b; Portmann et al.,
The inferior colliculus potential should thus be of 1980). However, in estimating the period of time
rather large amplitude, but also one which is rela- of continuing activity of the eighth nerve, animal
tively broad because of the staggered input. and human studies have arrived at different con-
clusions. In subhuman mammals, wave II appears
to be the first central evoked potential. Depth
 39

electrodes in the cochlear nuclei of the cat record the sense of causing positive or negative deflec-
activity with a latency comparable to that of wave tions in the far-field recordings. Buchwald’s com-
II, and a section of the eighth nerve at the level of prehensive review (1983) discusses the question of
the internal acoustic meatus abolishes wave II and the relative contribution of graded postsynaptic
all following waves (Buchwald and Huang, 1975; potentials versus all-or-none action potentials. She
Henry, 1979; Achor and Starr, 1980b). In the cat, concludes that neuronal postsynaptic depolariza-
wave II follows wave I by about 1 ms, a period of tion potentials contribute more to the extracellular
time which is reasonable for conduction time in a field potential, and assumes that this contribution
5 mm long cochlear nerve consisting predomi- would cause a positive deflection in the surface
nantly of 2-5 pm axons (Arnesen and Osen, 1978) record. There is, however, some evidence from
plus the synaptic delay in the cochlear nuclei. In recordings made from electrodes placed directly
man, anatomical studies indicate much a longer on the human brainstem during neurosurgery
cochlear nerve, one 25 mm in length (Lang, 1981) which indicates that activity in neural centers is
but consisting of axons very similar in diameter to reflected as a negative wave in the surface record-
that of the cat, i.e. 2-6 pm (Terr, Pers. Commun.). ing. In recording from electrodes placed on the
The data on length of the human cochlear nerve surface of the cochlear nuclei and inferior collicu-
strongly suggests that conduction time in the hu- lus, Moller and Jannetta (1981, 1982b,c) detected
man eighth nerve is longer than in animals, and a slow negative wave, which was interpreted as
this is supported by direct recordings from the being generated by depolarization of the neurons
nerve in neurosurgical patients. Both Moller in those structures. In both cases, the negative
(Moller et al., 1981; Moller and Jannetta, 1981, wave was preceded and followed by sharper posi-
1982b) and Hashimoto (Hashimoto et al., 1981) tive peaks, thought to have been generated in the
obtained recordings from electrodes on the in- afferent and efferent axonal pathways of the
tracranial portion of the human eighth nerve which nucleus.
show a triphasic potential with a positive peak Although there is at present no definitive answer
matching wave II in the scalp-recorded ABR. On to this question, the best fit of brain stem anatomy
the basis of these studies, the earliest evoked activ- to ABR waveform is obtained if the assumption is
ity in the human brainstem nuclei would not be made that somatodendritic potentials in cell groups
expected until sometime after wave II. are reflected as negative deflections and massed
On the basis of the anatomy of the human axonal activity in fiber tracts as positive waves.
central auditory pathway, one likely sequence of Using the terminology employed by Hashimoto et
central events set off by a volley of impulses in al. (1979, 1981) to indicate both the positive and
one cochlear nerve would be the following. A negative waves (Fig. 3A), the best fit of the known
synchronized somatodendritic depolarization in anatomy of the human brain stem pathway to the
the cochlear nuclear complex would be followed ABR can be obtained by the schema illustrated in
shortly by a depolarization in both medial olivary Fig. 3B. In this interpretation, the size, shape, and
nuclei. Somewhat later, a much smaller depolari- latencies of the negative waves are matched to the
zation would occur in the dorsal lemniscal nuclei, predictions made on the basis of relative size and
and shortly after that a large, but less well syn- position of the central nuclei.
chronized depolarization would occur in the infe- (1) Wave N2, the sharp and well defined wave
rior colliculus. During this time, action potentials immediately following the P2 potential generated
would be continuously present in both the direct by auditory nerve axons, would represent the first
and relayed pathways connecting the cochlear central response in the cochlear nuclei. This nar-
nuclei with the colliculus. The chief problem in row but high amplitude wave is what would be
matching even such a simplified schema of central expected of a large population of neurons firing in
activity with the actual ABR waveform is the relative synchrony.
question of how neuronal somatodendritic (2) Wave N3, a sharp negative peak, would
depolarization potentials and axonal action poten- reflect depolarization in the medial olivary nuclei.
tials are reflected in the scalp-recorded ABR in The peak of N3 follows that of N2 by about a
 the effectiveness of this structure as an generator.
(3) Wave N4, a small negative deflection which
creates waves IV and V, would represent depolari-
zation of the dorsal lemniscal nuclei. Wave Iv4 is a
shallow and somewhat less well defined potential.
It follows wave N2 by about 2 ms, a latency which
is reasonable for conduction in the 38 mm path-
way from the contralateral cochlear nuclei, plus a
synaptic delay in the dorsal nucleus. In addition,
it receives input from both medial olivary nuclei.
Lack of synchrony in afferent pathways to the
dorsal nucleus could account for the fact that this
wave is broad relative to its depth.
(4) Wave N5, the broad and deep negative
trough following wave V, would represent the
depolarization of neurons in the inferior collicu-
lus. Wave NS is the most pronounced negative
wave in the record and one which could reasona-
k bly be generated by the inferior colliculus, the
largest nuclear structure in the human brain stem
auditory pathway. The total period of negative
i ‘ ’
2 4 * s
1
(0 Inem activity reflected in wave N5 covers a period of
0

Fig. 3. (A) Positive and negative waves of the ABR according

almost 2 ms. A response with this spread in time
to the terminology of Hashimoto et al. (1979, 1981). (3) would be expected from the colliculus, given that
Proposed correlation of neural centers with the ABR wave- it should receive staggered input from the path-
form. Axonal potentials in the trapezoid body (tb) and faterai ways arising in the contralateral cochlear nuclei, in
temniscus (11) are thought to be reflected in positive waves.
both medial olivary nuclei, and in both dorsal
Depolarizations in nuclear centers such as the cocblear nuclei
(cn), medial olivary nuclei (mso), dorsal Iemniscal nuclei (dnll), lemniscal nuclei (Figs. 2A, B).
and inferior coliiculus (ic) are thought to create negative de The converse assumption, that the positive
flections. peaks represent axonal activity, would mean that
wave P2 (wave II) represents activity in the
cochlear nerve, wave P3 (wave III) reflects axons
millisecond, which would represent transmission firing chiefly in the trapezoid body, and the entire
time in trapezoid body axons, plus the time of a positive wave capped by peaks P4 and P5 (waves
synaptic delay in medial nucleus neurons. The IV and V) reflects axonal activity in the lateral
width of N3, indicating the total period of time in lemniscus. In this interpretation, action potentials
which a negative potential is present, is more than at the time of P3 would be traveling in a predomi-
a millisecond. This period is sufficient to allow for nantly horizontal course across the brain stem,
activity in both the ipsilateral and contralateral while axonal activity at the time of P4-PS would
medial nuclei, slightly separated in time in their be passing vertically along the long axis of the
response. Wave N3 is a large and consistent brain stem. Since both the cochlear nuclei and
potential of the type that would be expected from inferior colliculus are obligatory synapses in the
structures like the medial olivary nuclei, which act ascending auditory pathway, a cessation of axonal
as effective dipoles and produce a prominent activity should occur at the time of their response.
extracellular field potential. In addition, neurons This cessation of activity would occur within a
of the medial olivary nucleus have been shown to short time period at the level of the cochlear
have precise, invariant latencies and other ‘time- nuclei, but would be much less synchronous at the
keeping’ properties (Galambos et al., 1959; Gold- colliculus. Between these points, there would be
berg and Brown, 1969) which should contribute to continuous axonal action potentials traveling in
 41

the trapezoid body and lateral lemniscus, but at first evoked activity in the medial geniculate
the time of firing of the medial olivary and dorsal nucleus.
lemniscal nuclei, the far-field recording would re-
flect the algebraic sum of positive axonal poten- Clinical implications
tials and the negative somatodendritic depolariza-
tions in these nuclei. The issue of interpretation of the ABR in the
In addition to being a good fit to the structures case of pathological alteration is complicated by
of the human brain stem pathway, this hypothesis the fact that human auditory centers, and their
is in good accord with the results of direct human evoked responses, are neither totally independent
brain stem recording. Intraventricular electrodes nor totally interdependent. The medial olivary
(Hashimoto et al., 1981) recorded prominent P2, nucleus receives all of its ascending input from the
N2, P3, N3 and P4 waves from the caudal end of cochlear nuclei, the dorsal lemniscal nucleus is
the fourth ventricle, near the cochlear nuclei and innervated directly by the cochlear nuclei and also
olivary complex, but the record was dominated by indirectly through the medial olivary nuclei, and
a large P5-N5 complex when the electrode was the inferior colliculus receives axons from all three
positioned in the midbrain aqueduct close to the lower brainstem nuclei. Despite this complexity, it
inferior colliculus. Moller and Jannetta (1982a) should be possible to correlate the location of
recorded first negative, and then positive poten- pathological states such as tumors, infarcts, or
tials from the point of entry of the eighth nerve demyelinating lesions with changes in the ABR. If
into the brain stem, i.e., the cochlear nuclei, at a the model presented here is correct, lesions at the
time matching wave III of the scalp-recorded re- level of the pontomedullary junction, which have
sponse. These potentials presumably represent the the potential to involve the cochlear nuclei,
depolarization of cochlear nucleus neurons fol- trapezoid body axons, and the superior olivary
lowed by discharge of their efferent axons, and complex, would cause changes primarily in the
would correspond to peaks N2 and P3. With two N2-P3-N3 complex. The N2 peak, representing
electrodes located on the dorsal surface of the the cochlear nuclei, should be affected only by
human pons, Hashimoto et al. (1981) recorded lesions ipsilateral to the stimulated ear, but the
activity from a vertically oriented dipole with a P3-N3 peaks reflecting activity in the trapezoid
rostra1 positive and caudal negative potential in body and medial olivary nuclei could be affected
the time period of P4 and P5. This activity is most by ipsilateral, contralateral, or central brain stem
likely to be action potentials traveling rostralward lesions. Lesions in the middle to upper pons would
in the lateral lemniscus. In recordings from the alter or delay the P4-N4-P5 combination of peaks
surface of the inferior colliculus, Moller and Jan- because of involvement of the lemniscal pathway
netta (1982~) recorded a positive potential with a and possibly also of the dorsal lemniscal nuclei.
latency matching wave P5 (wave V), which they Finally, only lesions of the dorsal and posterior
believed to be generated by lemniscal axons. This midbrain should directly affect N5, the response
was followed by a broad negative potential coin- of the inferior colliculus, without involving some
ciding in time with the trough between peaks P5 of the preceding waves. In terms of laterality, the
(wave V) and P6 (wave VI). This N5 potential was lateral lemniscus would carry evoked activity on
interpreted as a depolarization response of neu- both sides of the brain, with contralateral activity
rons in the inferior colliculus. originating mainly in the cochlear nuclei and ipsi-
The information on brain stem morphology on lateral activity originating mainly in the medial
which this hypothesis is based (Moore, 1987a) olivary nucleus. Thus lesions on either side of the
does not, in itself, shed any light on the origin of upper pons could potentially affect the later waves
potentials subsequent to N5. Those positive waves of the ABR.
and the troughs separating them (P6, N6, P7, N7 The ultimate proof or disproof of this schema
and P8) might be produced by a later firing popu- of ABR generation can only be accomplished by
lation of neurons and axons in the inferior col- correlation with clinical cases and with electro-
liculus and its brachium, but might also reflect the physiological investigations, such as those employ-
42

ing three-dimensional trajectory analysis. Consid- tion of the dog superior olivary complex: an anatomical
eration of this model, in conjunction with future and el~trophysiolo~&al study. J. Neurophysiol. 31,
639-656.
clinical and experimental investigations, may con-
Goldberg, J.M. and Brown, P.B. (1969) Response of binaural
tribute to greater accuracy in use of the ABR in neurons of dog superior olivary complex to dichotic tonal
clinical testing, surgical monitoring, and patient stimuli: some physiological mechanisms of sound IocalizAt-
diagnosis. tion. J. Neurophysiol. 31, 613-636.
Goldberg, J.M. and Moore, R.Y. (1967) Ascending projecttons
of the lateral lemniscus in the cat and monkey. J. Comp.
Acknowledgements
Neurol. 129, 143-156.
Guinan, J.J.Jr., Norris, B.E. and Guinan, S.S. 11972) Single
The author is grateful to Dr. Nancy Squires for auditory units in the superior olivary complex. II. Location
discussion and critical comments. This work was of unit categories and tonotopic organization. Int. J. Neu-
supported by the Deafness Research Foundation. rosci. 4, 147-166.
Hall, J.G. (1965) The cochlea and cochlear nuclei in neonatal
asphysia. A histological study. Universitets Forlaget, Oslo.
References Harrison, J.M. and Irvin, R. (1966) Ascending connections of
the anterior ventral co&ear nucleus in the rat. J. Camp.
A&or, L.J. and Starr, A. (1980a) Auditory brainstem responses Neural. 126, 51-64.
in the cat. f. Intracranial and extracranial recordings. Eke- Hashimoto, I., Ishiyama, Y. and Tozuka, G. (1979) Bilaterally
troencephalogr. Clin. Neurophysiol. 48, 154- 173. recorded brainstem auditory evoked responses: their asym-
Achor, L.J. and Starr, A. (1980b) Auditory brain stem re- metric abnormalities and lesions of the brainstem. Arch.
sponses in the cat. II. Effects of lesions. Electroencepha- Neurol. 36, 161-167.
logr. Clin. Neurophysiol. 48, 174-190. Hashimoto, I., Ishiyama, Y. and Yoshimoto, T. (1981) Brains-
Adams, J.C. (1979) Ascending projections to the inferior coi- tern auditory evoked potentials recorded directly from hu-
liculus. J. Comp. Neurol. 183, 519-538. man brainstem and thalamus. Brain 104, 841-859.
Aitkin, L.M. and Phillips, SC. (1984) Is the inferior colliculus Hecox, K. and Galambos, R. (1974) Brainstem auditory evoked
an obligatory relay in the cat auditory system? Neurosci. responses in human infants and adults. Arch. Otolaryngol.
Lett. 44, 259-264. 99, 30-33.
Amesen, A.R. and Osen, K.K. (1978) The cochlear nerve in the Henkel, C.K. and Spangler, KM. (1983) Organization of the
cat: topography, cochleotopy, and fiber spectrum. J. Camp. efferent projections of the medial superior olivary nucleus
Neurol. 178, 661-678. in the cat as revealed by HRP and autoradiographic tracing
Beyerl, B.D. (1978) Afferent projections to the central nucleus methods. J. Comp. Neurol. 221, 416-428.
of the inferior collicuIus in the rat. Brain Res. 145,209-223. Henry, K.R. (1979) Auditory brainstem volume conducted
Brunso-Bechtold, J., Thompson, G.C. and Masterton, R.B. responses. Origins in the laboratory mouse. Am. J. Aud.
(1981) Study of the organization of auditory afferents Sot 4, 173-178.
ascending to the central nucleus of the inferior coIliculus in Huang, C-M. and Buchwald, J.B. (1977) Interpretation of the
the cat. 3. Comp. Neurol. 197, 705-722. vertex short-latency acoustic response: A study of single
Buchwald, J.S. (1983) Generators. In: Bases of Auditory neurons in the brain stem. Brain Res. 137,291-303.
Brain-Stem Evoked Responses, pp. 157-193. Editor: E.J. Kane, ES. and Barone, L.M. (1980) The dorsal nucieus of the
Moore. Grune and Stratton, New York. lateral lemniscus in the cat: Neuronal types and their
Buchwald. J.S. and Huang, C.-M. (1975) Far-field acoustic distributions. J. Comp. Neurol. 192, 797-826.
response: origins in the cat. Science 189, 382-384. Kiang, N.Y., Fullerton, B.C., Richter, E.A.. Levine, R.A. and
Clark, G.M. and Dunlop, C.W. (1968) Field potentials in the Norris, B.E. (1984) Artificial stimulation of the auditory
cat medial superior olivary nucleus. Exp. Neurol. 20, 31-42. system. Adv. Audiol. 1, 6-17.
Dublin, W.B. (1976) Fundamentals of Sensorineural Auditory Kudo, M. (1981) Projections of the nuclei of the lateral
Pathology. Charles C. Thomas, Springfield IL. lemniscus in the cat: an autoradiographic study. Brain Res.
Fernandez, C. and Karapas, F. (1967) The course and termina- 221, 57-70.
tion of the striae of Monakow and Held in the cat. J. Lang, J. (1981) Facial and vestibulocochlear nerves: topo-
Comp. Neurol. 131, 371-386. graphic anatomy and variations in: The Cranial Nerves, pp.
Galambos, R., Schwartzkopff, J. and Rupert, A. (1959) Micro- 363-377. Editors: M. Samii and P.J. Jannetta. Springer,
electrode study of superior olivary nuclei. Am. 3. Physiol. Berlin.
197, 521-536. Lev, A. and Sohmer, H. (1972) Sources of averaged neural
Geniec, P. and More% D.K. (1971) The neuronal architecture responses recorded in animal and human subjects during
of the human posterior colliculus: a study with the Golgi cochlear auditometry (electr~ochleogram). Arch. Ktin. Exp.
method. Acta Oto-Laryngol., Suppl 295, l-33. Ohren-, Nasen-, Kehlkopfbeilkd. 201, 79-90.
Glendenning, K.K., Brunso-Bechtold, J., Thompson, G.C. and Masterton, R.B. and Imig, T.I. (1984) Neural mechanisma for
Masterton, R.B. (1981) Ascending auditory afferents to the sound localization. Annu. Rev. Physiol. 46, 275-287.
nuclei of the lateral lemniscus. J. Comp. Neurol. 197. Meller, A.R. and Jannetta, P.J. 11981) Compound actton
673-704. potentials recorded intracranially from the auditory nerve
Goldberg, J.M. and Brown, P.B. (1968) Functional organiza- in man. Exp. Neurol. 74, 862-874.
 43

Moller, AR. and Jannetta, P.J. (1982a) Auditory evoked tials recorded from the external ear in man. Ann. Otol.
potentials recorded intracranially from the brainstem in Rhinol. Laryngol. 76, 427-435.
man. Exp. Neurol. 78, 144157. Starr. A. and Achor, L.J. (1975) Auditory brainstem responses
Molter. A.R. and Jannetta, P.J. (1982b) Comparison between in neurological disease. Arch. Neurol. 32, 761-76X.
intracranially recorded potentials from the human auditory Starr, A. and Hamilton, A.E. (1976) Correlation hetween con-
nerve and scalp recorded auditory brainstem responses. firmed sites of neurological lesions and abnormalities of
Stand. Audio]. 11. 33-40. far-field auditory brain stem responses. Electrocncephalogr.
Meller, A.R. and Jannetta, P-3. (198%) Evoked potentials from Clin. Neurophysiol. 41, 549-608.
the inferior colliculus in man. Electroencephalogr. Clin. Stockard. J.J. and Rossiter. VS. (1977) Clinical and pathologic
Neurophysiol. 53, 612-620. correlates of brain stem auditory response abnormalities.
Molfcr, A.R.. Jannetta, P.J., Bennett, M. and Meller. M.B. Neurology 27, 316-325.
(1981) Intracranially recorded responses from the human Stromingrr, N.L. (1973) The origins, course, and distribution
auditory nerve. New insights into the origin of brainstem of the dorsal and intermediate acoustic striae in the rhesus
evoked potentials (BSEPs) Electroencsphalogr. Clin. Neu- monkey. J. Comp. Neurol. 147. 209-234.
rophysiol. 52. 18-27. Strominger, N.L. and Hurwitz, J.L. (1976) Anatomical aspects
Moore. J.K. (1987a) The human auditory brainstem: A com- of the superior olivary complex. J. Comp. Neural. 170.
parative view. Hear. Res. 29, l-32. 485-498.
Moore, J.K. (1987b) The dorsal cochlear nucleus: efferent Strominger. N.L. and Strominger. A.I. (lQ71) Ascending
projections and intrinsic organization. In prep. brainstem projections of the anteroventral cochlear nucleus
Moore, J.K. and Moore. R.Y. (1971) A comparative study of in the rhesus monkey. J. Camp. Neural. 143, 217-242.
the superior olivary complex in the primate brain. Folia Strominger, N.L., Nelson, L.R. and Dougherty, N.J. (1977)
Primatol. 16, 35-51. Second order auditory pathways in the chimpanzee. J.
Moore, J.K. and Osen, K.K. (1979) The cochlear nuclei in Comp. Neural. 179, 349-366.
man. Am. J. Anat. 154, 393-418. Terr, L.I. and Edgerton, B.J. (1985) Surface topography of the
Mnrest, D.K. and Oliver, D.L. (1984) The neuronal architcc- cochlear nuclei in humans: two- and three-dimensional
ture of the inferior colliculus in the cat: defining the analysis. Hear. Res. 17. 51-59.
functional anatomy of the auditory midhrain. J. Comp. Tsuchitani, C. and Boudreau, C. (1964) Wave activity in the
Neural. 222, 209..-236. superior olivary complex of the cat. J. Neurophysiol. 27.
Nordeen. U.W., Kiliackey, H.P. and K&es, L.M. (1983) X14-827.
Ascending auditory projections to the inferior colliculus in Wada, S. and Starr, A. (1983a) Generation of auditory brains-
thr adult gerbil, Merro~res ungu~culafur. J. Comp. Neural. tern responses (ABRs): I. Effects of injection of a local
214. 131-143. anesthetic (procaine HCI) into the trapezoid body of guinea
0110. C. and Schwartz, I.R. (1979) The superior ohvary com- pigs and cat. El~tr~nccphalogr. Clin. Ncurophysiol. 56.
plex in C57BL/6 mice. Am. J. Anat. 155. 349-379. 326-339.
Osen, K.K. (1972) Projection of the cochlear nuclei on the Wada, S. and Starr, A. (1983b) Generation of auditory brains-
inferior colliculus rn the cat, .I. Camp. Neurol. 144. 355-372. tern responses (ABRs). II. Effects of surgical section of the
Picton. T.W., Hillyard, S.A., Krause, H.I. and Galamhos, R. trapezoid body on the ABR in guinea pigs and cat. Electro-
(1974) Human auditory evoked potentials. 1. Evaluation of encephalogr. Clin. Neurophysiol. 56, 340-351.
components. Electroencephalogr. Clin. Neurophysiol. 36, Wada, S. and Starr, A. (1983~) Generation of auditory brains-
176-190. tern responses (ABRs). III. Effects of lesions of the superior
Portman, M., Cazals. Y., Negrevergne, M. and Aran, J.M. olive. lateral lemniscus. and inferior colliculus on the ABR
(1980) Transtympanic and surface recordings in the diagno- in the guinea pig. Electroencephalogr. Clin. Neurophysiol.
sis of retrocochlear disorders. Acta Otolaryngol. (Stockh.) 56. 3522366.
X9, 362-369. Warr, W.B. (1966) Fiber degeneration following lesions in the
Ramoh Y Cajal, S. (1909) Histologie du Systeme Nerveux de anterior ventral cochlear nucleus of the cat. Exp. Neurol.
1’Homme et des Vertebres, pp. 774-838. Maloine, Paris. 14.453-474.
Ryugo. D.K.. Willard, F.H. and Fekete, D.M. (1981) Differen- Warr. W.B. (1969) Fiber degeneration following lesions in the
tial afferent projections to the inferior colliculus from the posteroventral cochlear nucleus of the cat. Exp. Neural. 23.
cochlear nucleus in the albino mouse. Brain Res. 210, 140-155.
3422349. Watr, W.B. (1982) Parallel ascending pathways from the
Scheibel, M.E. and Scheibel, A.B. (1974) Neuropil organization cochlear nucleus: Neuroanstomical evidence of functional
in the superior olive of the cat. Exp. Neural. 44, 339 348. specialization. Contrib. Sens. Physiol. 7, l-38.
Schwartz, 1.R. (1977) Dendritic arrangements in the cat medial Wcbstcr, D.B. (1971) Projection of the cochlea to cochlear
superior olive. Nrurosci. 2, 81-101, 1977. nuclei in Merriam’s kangaroo rat. J. Comp. Neural. 143.
Schweizer. H. (1981) The connections of the inferior colliculus 323-340.
and the organization of the brain stem auditory system in Zook. J.M. and J.H. Casseday (1980) Identification of auditory
the greater horseshoe bat (Rhinolophus ferrumequmum ). J. centers in lower brainstem of two species of echolocating
Camp. Neural. 201. 25-49. bats: evidence from injection of horseradish peroxidase
Shneiderman, A. and Henkel, C.K. (19X5) Evidence of col- into inferior colliculus. In: Proc. 5th In:. Bat Res. Conf..
lateral axonal projections to the superior olivary complex. pp. 51-60. Editors: D.E. Wilson and A.E. Gardner. Texas
Hear. Res. 19, 199-205. Tech Press. Lubbock, Texas.
Sohmer. H. and Feinmesser. M. (1967) Cochlear action poten-