The periaqueductal grey area is an area of neuron cell bodies that surrounds the cerebral aqueduct. It is continuous with the grey substance of the third ventricle.

The tectum comprises the ‘roof’ or dorsal portion of the midbrain and contains the corpora quadrigemina, which includes the superior and inferior colliculi and all the substances that lie dorsal to the cerebral aqueduct in the midbrain area of the neuraxis. The superior colliculi are involved with visual reflexes, and project to the lateral geniculate bodies of the thalamus. The inferior colliculi are involved with reflexes associated with sound, and project to the medial geniculate bodies of the thalamus.

The tegmentum consists of the bulk of the matter of the brainstem and comprises the area ventral to the cerebral aqueduct and fourth ventricle. It contains the bulk of the brainstem nuclei and the reticular formation of the midbrain, pons, and medulla.

The substantia nigra is a broad layer of pigmented neurons that separates the basis from the tegmentum. It extends from the upper surface of the pons to the hypothalamus. It projects to and receives projections from the neostriatum, thalamus, subthalamic nucleus, superior colliculi, and the reticular system of the brainstem. The neuron of this region utilise dopamine as a neurotransmitter.

The oculomotor nuclei are the motor nuclei of the superior rectus, inferior rectus, and medial rectus muscles of the eye.

These bilateral structures are ovoid groups of nuclei composed of two different cell types, the magnocellular and parvocellular groups of neurons. The magnocellular neurons are large, multipolar cells located in the caudal area of the red nuclear mass. These neurons receive bilateral projections both from sensorimotor cortical areas via the corticorubral tracts and from collaterals via the corticospinal tracts. The cortical projections and their target neurons in the red nucleus are somatotopically organised. Axon projections from the magnocellular neurons form the rubrospinal tracts, which cross in the brainstem and project in a somatotopically organised fashion, mainly to the interneurons of the intermediate grey areas of the spinal cord. Some rubrospinal fibres terminate directly on ventral horn motor neurons as well. Some axons that form the rubrospinal tracts terminate on neurons in the pontomedullary reticular formation and the motor nuclei of various cranial nerves, forming the rubroreticular system and the rubrobulbar tracts, respectively (Brown 1974). Reciprocal, bilateral projections to the superior colliculi are also present and form the rubrotectal tracts (Fig. 13.6). The rubrospinal and corticospinal tracts form the lateral motor system of the spinal cord. The medial motor system is composed of the reticulospinal and vestibulospinal tracts.

Figure 13.6 Afferent and efferent projections of the red nucleus.

The parvocellular neurons of the red nucleus are small pyramidal- and spherical-shaped neurons, mostly located in the rostral areas of the red nuclear mass. These neurons receive projections from the dentate nucleus of the contralateral cerebellum, and from the ipsilateral globus pallidus pars externa, substantia nigra, and subthalamic nuclei. These neurons project to the ipsilateral thalamus (Fig. 13.6).

This structure is a highly myelinated axon tract that descends from the interstitial nucleus of Cajal in the lateral wall of the third ventricle through the midbrain, pons, and medulla to the spinal cord where it becomes continuous with the anterior intersegmental fasciculus. The MLF acts as a major communications conduit between all of the cranial nerve nuclei and all related structures including the reticular formations of the mesencephalon, pons, and medulla. The medial vestibulospinal tract axons project with the MLF to the spinal cord (Brodal et al. 1962). Other axons from the vestibular nuclei ascend in the MLF to more rostral structures including the extraocular cranial nuclei. The MLF is the first fibre tract in the developing embryo to become myelinated and probably acts as a major pathway providing stimulus to developing neurons in the early stages of embryonic development.

The medial lemniscus is a bundle of axons that forms a triangular structure medial to the spinothalamic tract. The bundles are formed from axons of the contralateral dorsal column nuclei, the nucleus gracilis and cuneatus, which have decussated and formed the internal arcuate fibres which become continuous with the medial lemnisci. The fibres are joined by axons of the trigeminal sensory nucleus to project to the ipsilateral thalamus. Axons from the dorsal column nuclei terminate on neurons in the ventral posterior lateral nucleus of the thalamus, whereas axons from the trigeminal sensory nucleus terminate on neurons in the ventral posterior medial nucleus of the thalamus (Guyton & Hall 1996).

This semilunar structure, also referred to as the basis, is located anterior to the substantia nigra and is composed of the corticospinal, corticonuclear, and corticopontine fibre tracts. The corticospinal fibres terminate on the ventral horn neurons of the contralateral spinal cord. The corticonuclear fibres terminate mainly on contralateral cranial nerve nuclei throughout the brainstem. The corticopontine fibres are composed of projections from the frontal and temporal cortical areas and terminate on the interneurons of the nuclei pontis. These nuclei then project mostly to the contralateral cerebellum.

This structure, which is also referred to as the brachium conjunctivum, proceeds from the upper white substance of the cerebellar hemisphere to the tegmentum where it completely decussates at the level of the inferior colliculus. It is composed of:

These tracts form bilateral structures that ascend in the spinal cord in the ventral lateral fasciculi. They terminate in the vermis and intermediate zones of the ipsilateral cerebellum. These tracts relay information to the cerebellum about what information or commands have arrived at the ventral horn cells. These pathways are part of the efference copy mechanism of the cerebellar motor system.

This tract carries fibres from the contralateral dorsal cochlear nucleus to the inferior colliculus.

These bilateral structures, which are also referred to as the brachium ponti, are the largest of the cerebellar peduncles. They carry fibres from the pontine nuclei to the contralateral neocerebellum. They are a component of the corticopontocerebellar pathways.

The reticular formation of the pons is continuous with the reticular formations of the medulla and the mesencephalon. This area receives input from virtually all areas of the neuraxis and projects widely throughout the neuraxis. (See below for more detail.)

These structures were discussed earlier; see above.

This cavity is bounded by the pons and medulla ventrally and by the cerebellum dorsally. It is continuous with the cerebral aqueduct above and the central canal of the medulla below and has a capacity for cerebrospinal fluid (CSF) of about 20 ml. The floor of the fourth ventricle, which is also referred to as the rhomboid fossa, is formed by the dorsal surfaces of the pons and medulla. The fourth ventricle acts as a component of the CSF system (Chusid 1982).

Structures found at a cross-section at the inferior olive (fig. 13.8)

Choroid plexus of 4th

This structure is also referred to as the tela choroidea of the fourth ventricle and is composed of a layer of pia matter that has become highly vascularised. The choroid plexus produces CSF.

Figure 13.8 A cross-sectional view of the medulla at the level of the inferior olive.

(from Standring 2008 Gray’s Anatomy 40th edn. Edinburgh, Elsevier, with permission)

Inferior olivary nucleus

These bilateral groups of nuclei are located within the olive of the medulla. They receive projections from the cortex, from other brainstem nuclei, from the ipsilateral parvocellular red nucleus, and from the ipsilateral spinal cord. These neurons project axons, referred to as climbing fibres, via the inferior cerebellar peduncles to all areas of the cerebellar cortex. These structures are part of a group of projections and nuclei referred to as the inferior olivary nuclear complex, which forms a complicated loop from the cerebellar cortex to the dentate nucleus to the contralateral red nucleus to the inferior olivary nucleus and back to the contralateral cerebellar cortex (Fig. 13.9)

Figure 13.9 The functional loop of the inferior olivary complex.

Structures found in a cross-section at the lemniscal decussation (fig. 13.10)

Fasciculus gracilis

These bilateral tracts run dorsally in the medulla medial to the cuneate fasciculi. They are formed by the axons of the dorsal root ganglion cells that detect proprioception and touch in the lower limbs and trunk, usually below the T6 level. The axons enter the spinal cord via the dorsal root and ascend ipsilaterally in the fasciculus gracilis to the gracile nucleus.

Figure 13.10 A cross-sectional view of the medulla at the level of the lemniscal decussation.

(from Standring 2008 Gray’s Anatomy 40th edn. Edinburgh, Elsevier, with permission)

Structures found at a cross-section through pyramidal decussation (fig. 13.11)

Fasciculus gracilis and cuneatus

This structure was discussed earlier; see above.

Figure 13.11 A cross-sectional view of the medulla through pyramidal decussation.

(from Standring 2008 Gray’s Anatomy 40th edn. Edinburgh, Elsevier, with permission)

In order to understand oculomotor control it is necessary to remember that all eye movements are designed to keep a desired object centred on the fovea, which allows for the greatest visual acuity. In order for the desired object to be clearly visualised, it must be held relatively steady on the fovea, and the two eyes must be simultaneously aligned to within a few minutes of arc (Leigh & Zee 1992). Understanding normal function allows us to have a better understanding of when and why abnormal eye movements occur. The normal tendency of the eyeball is to return to primary position. To hold the eyeball in any other position requires constant contraction of the extraocular muscles in exactly the right proportions.

When the eye moves to a new target it does so by a movement called a saccade, which is a fast, burst-like movement. Saccades can reach velocities of 700° per second and vision is transiently suppressed during saccadic movements. The saccade is programmed with two distinct components, a pulse phase and a step phase. The pulse phase is the burst of action potential activity to move the eye to the new target. The step phase is the new action potential firing rate to maintain the eye in the new position (Fig. 13.23).

Figure 13.23 Nerve activation phases of eye movements.

Saccades can be horizontal or vertical in nature. The burst phase of activity for a horizontal saccade is generated by burst neurons in the pontine paramedian reticular formation. The duration of firing of a burst neuron begins just before the saccade and ends exactly when the saccade enters the step phase. In between burst outputs, the burst neurons are tonically inhibited by omnipause neurons in the nucleus raphe interpositus (Buttner-Ennever & Buttner 1988). The omnipause neurons continuously discharge, inhibiting the burst neurons until they enter a pause cycle in which the burst neurons become disinhibited and fire a burst of action potentials that results in a saccade motion of the eye until the pause neurons resume their firing and inhibit the burst neurons. The step phase of the horizontal saccade is thought to be created by a neural gaze maintenance network or neural integrator that calculates the saccadic velocity to produce the appropriate position and to produce the appropriate contraction in the extraocular muscles to maintain the gaze at a specific point. The medial vestibular nucleus, the flocculus of the cerebellum, and the nucleus prepositus hypoglossi are important components of the neural integration system of horizontal movements (Zee et al. 1981; Cannon & Robinson 1987) (Fig. 13.24).

Figure 13.24 The components involved in the generation of horizontal eye movements.

The burst phase of activity for a vertical saccade is generated by burst neurons in the rostral interstitial nucleus of the medial longitudinal fasciculus (riMLF) (Buttner-Ennever & Buttner 1988). This nucleus is located ventral to the aqueduct of Sylvius in the prerubral fields at the junction of the mesencephalon and the thalamus. The activity in this nucleus is dependent on ascending projections from the pontine paramedian reticular formation, omnipause neurons in the nucleus raphe interpositus, and inputs from the contralateral vestibular nuclei. The riMLF projects to the ipsilateral oculomotor (CN III) and trochlear (CN IV) nuclei. Reciprocal connections occur between the right and left riMLF through the posterior and anterior commissures of the midbrain. The fibres of the elevator nuclei which innervate the superior rectus and the inferior oblique muscles pass through the posterior commissural pathways, and the projections to the depressor nuclei innervating the inferior rectus and superior oblique muscles pass through the anterior commissural pathways. The velocity commands of vertical saccadic movements are integrated by the interstitial nucleus of Cajal (Fukushima et al. 1990) (Fig. 13.25).

Figure 13.25 The components involved in the generation of vertical eye movements.

Cortical modulation of saccadic eye movements is also a necessary component for normal vision. The frontal lobe contains three areas that contribute to saccadic control; these are the frontal eye fields (FEF), the supplementary eye fields, and the dorsal lateral prefrontal cortex. The FEF are composed of a group of neurons in the posterior areas of Brodmann area 8 that discharge prior to saccadic movement and are thought to play a part in the initiation of saccades to previously seen or remembered targets (Bruce & Goldberg 1985). Excitation of these neurons results in contralateral saccades. Neurons in the supplementary FEF which are located in the dorsomedial frontal lobes are involved in programming saccades as part of complex learned behaviours (Mann et al. 1988). The dorsal lateral prefrontal cortex is involved with programming saccades to remembered target locations (Boch & Goldberg 1989).

Smooth pursuit movements allow viewing of moving objects. They are much slower than saccades and can reach a maximum velocity of only 100° per second. They are not under voluntary control. Smooth pursuit activities seem to be modulated in the middle and superior medial temporal visual areas. These areas of cortex project to the dorsal lateral median pontine nuclei (Suzuki et al. 1990). The dorsal lateral median pontine nuclei project to the flocculus, uvula, and dorsal of the cerebellum where the smoothness of the motion is calculated and adjusted.

The cerebellum is involved in two basic operations involving eye control. The first involves its role in real-time positional eye control with respect to visual acquisition, and the second involves long-term adaptive control mechanisms regulating the oculomotor system (Leigh & Zee 1991).

The cerebellum functions to ensure that the movements of the eyes are appropriate for the stimulation that they are receiving. The flocculus of the vestibulocerebellum contains Purkinje cells that discharge in relation to the velocity of eye movements during smooth pursuit tracking, with the head either stationary or moving. For example, you can keep your head still and fixate your gaze on a moving object, in which case your eyes should smoothly follow the object across your visual field. Or you could keep your eyes fixed on a stationary object and rotate your head, in which case your eyes should still smoothly track in the opposite direction and at the same speed as the rotation of your head to maintain the target in focus (Zee et al. 1981). Other neurons discharge during saccadic eye movement in relation to the position of the eye in the orbit. Individual control of eye movement is accomplished for the most part by the contralateral cerebellum although intimate bilateral integration is also important. For example, the smoothness of pursuit activity and the return to centre function of saccadic movement of the right eye are under left cerebellar modulatory control.

Disorders of saccadic eye movements can involve the accuracy, velocity, latency, and stability of the eye movements.

Disorders of smooth pursuit are often associated with dysfunction of the cerebellum or brainstem. Slow jerky pursuit motion can be associated with physiological decreases in activation due to medication or decreased cerebellar stimulation.

Nystagmus is a repetitive to and fro movement of the eyes that can be generated by watching moving targets such as occurs when looking out the window of a moving car or by moving an object such as an opticokinetic tape past an individual’s eye. When nystagmus occurs inappropriately it reflects dysfunction in the mechanisms that hold targets steady on the fovea of the retina. These mechanisms involve areas of the vestibular system, the neural integrator, and pursuit control systems. A variety of nystagmic patterns can be observed and related to specific dysfunctional areas.

The cochlea consists of three fluid-filled tubes in helical arrangement called the scala media, scala tympani, and scala vestibuli. The conduction of sound pressure waves occurs as the stapes, which acts as a piston on the round window, depresses the cochlear partition, which increases pressure in scala tympani. This increase in pressure results in an outward bowing of the round window. Any up and down motion of fluid is detected by the basilar membrane between the scala tympani and media.

The organ of Corti contains 16 000 hairs cells in four rows arranged as one inner row and three outer rows. The hair cells project into the gelatinous tectorial membrane and together support approximately 30 000 afferent nerve fibres and numerous efferent fibre terminals. Movement of the hair cells towards the tall edge results in excitation, whereas movement away from the tall edge or downward deflection results in inhibition.

The stapes footplate acts as a piston that pushes and pulls, causing a conduction of pressure waves through the fluid of the scala vestibuli. The pressure waves depress the cochlear partition, causing increase in pressure in the scala tympani and outward bowing of the round window. Up and down motion of the fluid is transmitted to the basilar membrane, resulting in deflection of the stereocilia on hair cells. Ca⁺⁺ and K⁺ enter through channels at the tips of stereocilia (‘gating springs’ or ‘tip links’), resulting in a receptor potential. Hair cells of vestibular and cochlear apparatus transduce sound and accelerations into electrical responses. They act as synaptic terminals by releasing chemical neurotransmitters to activate nerve fibre terminals when ion channels are opened by mechanical bending of stereocilia. The afferent fibres encode intensity, time course, and frequency.

The ventral cochlear nucleus is composed of two main cell types, the stellate cells, which are also known as chopper cells and give a steady, regular rhythm of impulses denoting stimulus frequency, and the bushy cells, which generate only one action potential and signify the onset of sound. Therefore, they provide accurate information about the timing of acoustic stimuli and are involved in locating sound stimuli along the horizontal axis (azimuthal).

The various cell types of the cochlear nuclei project along parallel pathways to specific relay nuclei that serve a common purpose. The VCN comprises two main divisions which include:

The DCN may be an important site of early auditory processing implicated in the physiology of tinnitus. In laboratory animals, the DCN has been found to comprise the following three layers that are parallel to the free surface of the brainstem:

The first and second layers are sometimes referred to as the superficial layers of the DCN. The molecular layer consists predominately of parallel fibres formed by the axons of granule cells and inhibitory interneurons including cartwheel and stellate cells. The anatomic organisation of the superficial layers of the DCN is therefore considered to be similar in many ways to that of the cerebellar folium.

The superficial layers of the DCN receive both auditory and nonauditory information including vestibular afferents, which are primarily from the saccule and somatosensory inputs. Within the superficial layer, the granule cells form excitatory connections with type IV units and inhibitory interneurons (especially cartwheel cells). In turn, the cartwheel cells form inhibitory connections on the type IV units. Type IV units are the output cells of the DCN and project to other components of the extralemniscal pathway as well as some neurons within the lemniscal pathway. Increased FOF of type IV units has been associated with the expression of tinnitus episodes and these neurons are exquisitely sensitive to sound. In the deeper layers of the DCN there are two inhibitory interneuronal circuits that have been identified:

Type II units have very low spontaneous rates of firing and give weak responses to broadband noise. They primarily respond to best frequency tones and provide inhibition to type IV units of the DCN. They are also thought to form inhibitory connections with bushy and multipolar cells of the VCN.

This nuclear area localises sound sources along the horizontal axis by distinguishing interaural time delays as small as 10 μs, and hence location to a few degrees.

A source in the midsagittal plane should excite two ears at the same time. Axon terminals from the contralateral anterior ventral cochlear nucleus (AVCN) excite successive neurons throughout the medial superior olive. Input from one ear is insufficient to bring an MSO neuron to threshold, so the interaural time difference is exactly counterbalanced by delay in conduction from the opposite ear.

Therefore, the simultaneous excitatory sound potential brings an MSO neuron to threshold and a map of sound source location along the horizontal axis is formed over an array of MSO neurons.

This area is also involved in localisation of sound but employs intensity cues rather than interaural time differences to calculate where a sound originated.

The LSO receives input from both cochlear nuclei such that ipsilateral inputs are direct and contralateral inputs from the sound source are via the nucleus of the trapezoid body. These inputs are antagonistic.

A given neuron in the LSO responds best when the intensity of a stimulus reaching one ear exceeds that on the opposite ear by a particular amount. The lateral olive is more efficient at processing high-frequency sounds because the head absorbs short wavelengths better than long wavelengths. This allows for clearer interaural intensity differences. Low-frequency sounds are processed more efficiently by the medial olive.

The lateral lemniscus includes axons from the superior olivary nucleus and the contralateral DCN via the dorsal acoustic stria. It provides passage for neurons to the inferior colliculus. The inferior colliculus consists of a dorsal region that receives both auditory and somatosensory inputs, and a central nucleus that comprises several layers forming a tonotopic map.

The multimodal division is sometimes referred to as the external nucleus of the inferior colliculus in research papers and forms part of the extralemniscal pathway.

The central nucleus of the inferior colliculus projects by way of the brachium of the inferior colliculus to the principal nucleus of the MGN. The principal nucleus of the MGN projects its neurons to the primary auditory area (A1 or 41, 42) on the transverse gyri of Heschl. The remaining components of the MGN are multimodal and form part of the extralemniscal pathway.

Consider the role of either hemisphere in the processing of auditory inputs from either field of space. This is analogous to the independent ocular dominance columns of the visual system.

Tinnitus has been referred to as a phantom auditory sensation that may occur because of reorganisation of the auditory cortex following some degree of deafferentation from the cochlea of the inner ear. Cochlear lesions resulting in loss of a specific range of frequencies can lead to reorganisation of the auditory cortex due to replacement of the corresponding cortical areas with neighbouring areas of sound representation (McIntosh & Gonzalez-Lima 1998).