Neural Basis Of Hallucinations Research Paper

Visual hallucinations: Peduncular hallucinations consist of animals, face, or human figures and are traditionally associated with vascular lesions in the upper brainstem (involving, especially, the cerebral peduncles). They can also occur in cases of encephalitis or during the transition from waking to sleep (Lhermitte 1922, from Brown 1985). Charles Bonnet syndrome is similar and occurs in the elderly, generally with significant visual impairment. However, human figures are characteristically small and doll-like, and can actually amuse the subject (Benson and Gorman 1993). Palinopsia is a pathological persistence of a visual percept originating from external stimulation that arises from lesions of the occipital cortex, seizures, or drug-induced states (Benson and Gorman 1993, Brown 1985). So-called elementary hallucinations consist of bands of light or phosphenes that can arise from pathology of the retina or optic nerve, and certain drug-induced states (Brown 1985). Hypnogagic hallucinations occur normally during the transition from sleep to wakefulness and are experienced as the continuation of a typical dreamstate. Hallucinations of faces or human figures are reported by approximately 20 percent of patients with schizophrenia. Visual hallucinations also often occur during seizures in patients with epilepsy. These hallucinations can have a geometric design or reflect a recollection of visual images based on past experience. Severe alcohol withdrawal often is accompanied by visual hallucinations of bugs or spiders. Lysergic acid (LSD) and related compounds often produce visual hallucinations consisting of light flashes, colors, geometric figures, and generally are accompanied by distortions of the geometric form of percepts (Crowley 1995). Higher doses produce fantastic forms, distorted facial images, etc. So-called autoscopic hallucinations reproduce an actual visual image of the self and are generally associated with epilepsy though these experiences also can occur in patients with schizophrenia or depression (Dening and Berrios 1994).

Auditory hallucinations: Most commonly, auditory hallucinations consist of spoken speech. These hallucinations are reported by 70–80 percent of patients with schizophrenia and less commonly by patients with other psychiatric disorders, such as depression or mania. These patients may also experience hallucinations of non-speech sounds, such as rattling, footsteps, etc. Other common causes of auditory hallucinations of speech include chronic alcoholism and phencyclidine use, or chronic amphetamine abuse. Hallucinations of speech consist of one or more ‘voices’ which comment on the patient’s speech, thought, or behavior, command them to perform certain actions, or converse with each other. A nonpathological variant of this experience is the momentary experience of hearing one’s name being spoken in the absence of an actual speaker. Both pathological and nonpathological variants of hallucinated speech are often triggered by indistinct sounds such as background conversation, street sounds, and running water. Often, if visual hallucinations are also experienced by patients with schizophrenia, they are phenomenologically linked with the auditory hallucination, e.g., a face will appear to speak the hallucinated words. Auditory hallucinations of music can occur in psychiatric patients with an affective disorder such as depression (Wengel et al. 1989) as well as in patients with deafness and steroid-induced psychotic states. Palinacousis is the auditory analogue of palinopsia and consists of the persistence of previously heard sounds, music, or speech.

Tactile or haptic hallucinations: These experiences commonly occur during the delirium tremens stage of severe alcohol withdrawal and consist of formication, i.e., the sensation of insects crawling on the body. Amphetamine or cocaine-induced psychosis can also lead to formication experiences. Phantom limb hallucinations follow arm or leg amputation and consist of the persistent sensation of the lost limb that can include the experience of movement and/or pain. Sensation of altered or duplicated body shapes can

occur in epilepsy, toxic encephalopathy, or with hallucinogenic drugs. When such hallucinations are accompanied by delusions that are fantastic, they are likely to be related to a schizophrenic process (Benson and Gorman 1993).

Hallucinations of bodily sense and movement have also been reported following direct electrical stimulation of the temporal lobe in humans (Ishibashi et al. 1964).

Olfactory hallucinations: These hallucinations most typically occur during seizure auras, migraine attacks, or as a result of damage to the olfactory bulb at the base of frontal lobe (e.g., by tumors). Olfactory hallucinations generally have a noxious content, consisting of a burning smell or some other foul odor. They are also reported occasionally by patients with schizophrenia and depression.

Mechanisms of hallucinations: Benson and Gorman (1993) outline a simple, but intuitively appealing, approach dividing hallucinations into two categories: the first ‘release’ hallucinations and the second ‘ictal’ hallucinations. The former arise when ‘normal sensory processing channels are blocked, allowing stored images to be experienced’ (pp. 266–267). Typical examples would be musical hallucinations following deafness or the Charles Bonnet syndrome arising from visual impairment in the elderly. These hallucinations tend to be longer in duration, can often be well-formed and quite elaborate, and occur only in the defective receptive field. ‘Ictal’ hallucinations arise from ‘abnormal neuronal discharge’ and tend to be briefer, less well-formed, and associated with alterations in consciousness. Benson and Gorman also highlight the importance of alterations in attention arousal in producing certain types of hallucinatory phenomena (1993, p. 278). Examples include dream-like hallucinations arising from the transition from sleep to arousal, the disturbed consciousness of drug-induced states, sleep deprivation, narcolepsy, or delirium. The postulated mechanism is an alteration in the gating of material into conscious awareness. In addition, limbic disorders and frontal dysfunction were postulated to cause hallucinations. We will discuss below some evidence suggesting a role of the former.

More recent work on the details of hallucinatory phenomena, considered in the remainder of the paper, suggests some revision may be necessary in the useful framework introduced by Benson and Gorman.

First, consider again phantom limb phenomena, a classic example of ‘release’ hallucination. These experiences are almost universal, occurring in approximately 85 percent of amputees (Benson and Gorman 1993) and are experienced as if the cortical representation of the limb is retained somehow as a conscious experience. Phantom limb experiences, especially early on, often are finely articulated with detailed sensations of the digits of the hands or feet. After loss of a hand, the person may still, for instance, re-experience the sensation of a long-worn ring. One puzzling observation is that patients who also lose sensation and use of a limb due to spinal cord injuries do not generally have phantom limb hallucinations or, if they do, these are very indistinct and fleeting. Why does loss of limb produce these experiences, while paraplegia does not? In both cases, deafferentation has occurred and ‘release’ phenomena should, in theory, arise.

Recent work by Spitzer et al. (1995) has provided an updated model suggesting that phantom limb hallucinations arise from a progressive reorganization of the sensory area of the cerebral cortex responsible for limb representation. The critical observation under- lying Spitzer’s model is that sensory ganglia in the spinal cord are known to fire randomly following denervation. These random impulses are then projected in a sustained fashion to the somatosensory cortex. Random firing from these sensory ganglia is not generated, in contrast, in the case of spinal cord injury. Using computer simulations, Spitzer demonstrated that chronic random inputs cause a deafferentized cortical sensory system to reorganize so that activation patterns becomes self-sustaining. More- over, a cortical region whose input is limited to noise was shown to assimilate receptive field properties of neighboring cortical areas gradually. Consequently, representations of the original limb representation are gradually usurped. Progressive reorganization of limb cortical maps in this fashion could produce the characteristic subjective ‘shrinking’ of these hallucinations over time—the phantom limb grows smaller over a period of weeks to months so that, in later stages, it is only experienced as, say, the size of a postage stamp superimposed on the stump.

The limits of the Benson Gorman approach are also suggested by the hallucinations experienced during amphetamine psychosis. These hallucinations typically consist of spoken speech of whole phrases or sentences or the perception of faces or bodily forms and are often accompanied by paranoid fears. Their phenomenology often resembles that of schizophrenia. Consequently, amphetamine psychosis is studied as an experimental model of this disorder.

These hallucinations do not readily fit a reduced attention/arousal model of hallucinations since amphetamine generally heightens attention and arousal. There are other data to suggest that perhaps amphetamine-induced hallucinations reflect an ‘ictal’ or ‘neuronal activation’ model of hallucinations. It is well-known, for instance, that a single dose of amphetamine produces widespread, heightened activation of the cerebral cortex (Porrino et al. 1984). However, amphetamine-induced hallucinations almost invariably do not emerge unless the person is a chronic abuser of the drug (Bell 1973).

Induction of apparent hallucinatory behavior in monkeys also requires repeated exposure to this drug (Castner and Goldman-Rakic 1999). Thus, it seems clear that the hallucinations induced by amphetamine do not arise simply from heightened brain activation—which is readily detected following a single exposure to the drug—but instead reflect complex changes in brain function generated over time by repeated drug exposure. The exact nature of these brain ‘sensitization’ changes has yet to be determined. However, recent studies suggest that alterations of neural circuitry in brain regions which receive projections from dopamine neurons and/ordinarily respond to amphetamine are critical (Pierce and Kalivas 1997). These alterations probably involve a cascade of molecular events that enhance cellular aspects of learning/conditioning mediated by receptors for the excitatory amino acid neurotransmitter, glutamate, as well as elaboration of new neuroanatomic connections (Robinson and Kolb 1999, Wolf 1998). These changes produce widely distributed neurocircuits that consequently become ‘primed’ to respond to the activating effects of amphetamine.

Reports of direct electrical stimulation in the human brain also suggest that neural activation alone is not sufficient to cause hallucinations. These studies have been conducted in patients with intractable epilepsy who were evaluated for neurosurgical resection of seizure loci. Penfield, a pioneer in this area, reported that complex hallucinations whose content consisted of ‘engrams’ of previous memories could be elicited in unanesthetized human subjects by electrically stimulating temporal lobe neocortical gray matter (cf. Penfield and Perot 1963). Penfield proposed that such stimulation directly accessed neural networks that actually coded for memories. Later workers found it difficult to obtain similar results, however. Halgren et al. (1978), Gloor et al. (1982), and Horowitz and Adams (1970), for instance, were able to elicit experiential responses but by stimulating deeper limbic structures (i.e., hippocampus and amygdala) rather than neocortical areas.

These experiences tended not to be memories but did include a range of hallucination-like experiences including unformed sensory experiences (lights, streaks, flashes) and complex visual experiences of persons, scenes, or objects or autoscopic experiences of the self. The location of sites that elicited hallucinations is curious since it is generally assumed that only neocortical structures (not limbic structures) code for actual sensory experiences. This raises the question of how hippocampal and amygdala stimulation, whose known functions pertain to registration of emotional salience, memory storage, and memory reactivation, could produce hallucinations.

One clue was that hallucinations were more likely to emerge when limbic stimulation produced not simply activation, but sustained afterdischarges, i.e., rhythmic activity detected via EEG recordings (Halgren et al. 1978, Horowitz and Adams 1970). After-discharges of this sort are more likely to lead to a temporary functional lesion of neural systems rather than activation of coded information (Gloor et al. 1982). Analogously, rhythmic electrical stimulation of neocortical tissue produces temporary aphasia or speech arrest if directed to language processing areas of the brain (Ojemann and Mateer 1979). Halgren et al. 1978) and Horowitz and Adams 1970) proposed that sustained disruption of limbic structures induced by after-discharges lead to the ‘release’ of sensory material from neocortical regions into consciousness. Consistent with this view are EEG studies in animals showing that the REM stage of sleep and LSD administration, which each produce vivid hallucinations in humans, cause dramatic reductions in hippocampal activation (Halgren et al. 1978). These observations suggest that the actual production of hallucinations by intracerebral stimulation arises from a dynamic interaction between neocortical and limbic systems, rather than simple activation of a group of neurons.

As with other forms of hallucinations, a new picture of auditory hallucinations in patients with schizophrenia is now emerging. Functional neuro-imaging studies indicate that these hallucinations are accompanied by activation from speech perception and auditory neurocircuitry (Silbersweig et al. 1995, Dierks et al. 1999); one of these studies also strongly implicated activation of hippocampal structures as well (Silbersweig et al. 1995). However, it is not clear if this distributed pattern of activation produces auditory hallucinations or arises as a consequence of some other pathophysiological process. For instance, shifts in attention due to the experience of hallucinations (from outward focus to internal imagery focus) may produce regional brain activation effects. If so, brain activation of this type will accompany hallucinations but will not in itself signal a causal mechanism.

Transcranial magnetic stimulation (TMS) is a method whereby an electromagnet placed on a scalp transmits magnetic pulses or waves to a small portion of the brain. Repetitive transcranial magnetic stimulation (rTMS) delivered at a low frequency (once per second) has been shown to reduce the reactivity or excitability of the part of the brain stimulated and other brain regions functionally connected to the region stimulated (Chen et al. 1997, Wassermann et al. 1998). A pilot study using a double-blind, crossover design compared effects of active versus sham low frequency rTMS in schizophrenic patients with auditory hallucinations (Hoffman et al. 1999, 2000). Based on the aforementioned neuro-imaging studies, the left temporoparietal cortex, a brain region critical for speech perception, was stimulated. A total of forty minutes of rTMS was administered over a four-day period. Statistically robust improvements in auditory hallucinations following active rTMS relative to sham rTMS were observed. These rTMS data support the hypothesis that activation of speech processing neurocircuitry contributes to the production of these hallucinations rather than emerging as a ‘downstream’ consequence of some other pathophysiological process.

These findings still do not address why speech processing areas of the brain, such as temporoparietal cortex, are pathologically activated in patients with schizophrenia. To understand this process better, computer simulations of neural networks have been developed that reproduce some aspects of speech processing of continuous narrative speech (Hoffman and McGlashan 1997). These simulations suggest that the ultimate ‘cause’ of auditory hallucinations is a loss of neuronal connections within these networks. Reductions in cortico-cortical connectivity are a normal component of ‘late’ development in human association cortex that continues during human adolescence (Huttenlocher and Dabholkar 1997). There are empirical data suggesting that reductions in synaptic density accompany some forms of learning (Scheich et al. 1991). Along these lines, computer simulations demonstrate that selective pruning of connections within recurrent neural networks produces significant information processing advantages (Hoffman and McGlashan 1997). In this model, pruning connections caused the system to fill in some of the gaps due to ambiguous phonetic input. It is well known that ordinary continuous speech, when produced at normal rates, has significant phonetic ambiguity that we do not ‘hear’ because our brains are also very skillful in ‘filling in the gaps.’ When pruning was excessive in our simulated networks, however, the system began to produce outputs in the absence of any phonetic input at all, thereby simulating speech hallucinations. This model therefore suggests that the hallucinations of schizophrenic psychosis arise as a result of an otherwise normal neuro-developmental trend that is excessive or unregulated. An appeal of this model is that it provides an account for the characteristic age of onset of schizophrenia, namely late adolescence and young adulthood.

Overall, the ‘release’ and ‘ictal’ models of hallucination described by Benson and Gorman appear to be good starting points. A comprehensive understanding of their nature, however, needs to consider interactions between different brain regions as well as between different ‘levels’ of neural processes involving molecular and neuroplastic alterations on the one hand, and large-scale neural networks dynamics on the other. In spite of our current limited understanding of hallucinations, this fascinating area is likely to provide important insights into the pathophysiology of brain disturbances, as well as the normal physiology of perception and consciousness.

Bibliography:

1. Bell D S 1973 The experimental reproduction of amphetamine psychosis. Archives of General Psychiatry 29: 35–40
2. Benson D F, Gorman D G 1993 The neural basis of hallucinations. Current Neurology, Vol. 13, Mosby-Year Books, pp. 265–87
3. Brown J W 1985 Hallucinations: Imagery and the microstructure of perception. In: Fredriks J A M (ed.) Handbook of Clinical Neurology. Elsevier, New York, Vol. 1, pp. 351–372
4. Castner S A, Goldman-Rakic P S 1999 Long-lasting pscyhotomimetic consequences of repeated low-dose amphetamine exposure in rhesus monkeys. Neuropsychopharmacology 20: 10–28
5. Chen R, Classen J, Gerloff C, Celnik P, Wassermann E M, Hallett M 1997 Depression of motor cortex excitability by low-frequency transcranial magnetic stimulation. Neurology 48: 1398–403
6. Crowley T J 1995 Hallucinogen-related disorders. In: Kaplan H I, Sadock B J (eds.) Comprehensi e Textbook of Psychiatry. Williams and Wilkins, Baltimore, MD, Vol. 1, pp. 831–8
7. Dening T R, Berrios G E 1994 Autoscopic phenomena. British Journal of Psychiatry 165: 808–17
8. Dierks T, Linden D E J, Jandl M, Formisano E, Goebel R, Lanfermann H, Singer W 1999 Activation of Heschl’s gyrus during auditory hallucinations. Neuron 22: 615–21
9. Foulkes D, Fleisher S 1975 Mental activity in relaxed wakefulness. Journal of Abnormal Psychology 84: 66–75
10. Gloor P, Olivier A, Quesney L, Andermann F, Horowitz S 1982 The role of the limbic system in experiential phenomena of temporal lobe epilepsy. Annals of Neurology 12: 129–44
11. Halgren E, Walter R D, Cherlow D G, Crandall P H 1978 Mental phenomena evoked by electrical stimulation of the human hippocampal formation and amygdala. Brain 101: 83–117
12. Hoffman R E, Boutros N N, Berman R M, Roessler E, Krystal J H, Charney D S 1999 Transcranial magnetic stimulation of left temporoparietal cortex in three patients reporting hallucinated ‘voices.’ Biological Psychiatry 46: 130–2
13. Hoffman R E, Boutros N N, Hu S, Berman R M, Krystal J H, Charney D S 2000 Transcranial magnetic stimulation of left temporoparietal cortex in schizophrenic patients reporting auditory hallucinations. Lancet 355: 1073–5
14. Hoffman R E, McGlashan T H 1997 Synaptic elimination, neurodevelopment and the mechanism of hallucinated ‘voices’ in schizophrenia. American Journal of Psychiatry 154: 1683–9
15. Horowitz M, Adams J 1970 Hallucination on brain stimulation: Evidence for revision of the Penfield Hypothesis. In: Kemp W (ed.) Origin and Mechanism of Hallucinations. Plenum Press, New York
16. Huttenlocher P R, Dabholkar A S 1997 Regional differences in synaptogenesis in human cerebral cortex. Journal of Comparative Neurology 387: 167–78
17. Ishibashi T, Hori H, Endo K, Sato T 1964 Hallucinations produced by electrical stimulation of the temporal lobes in schizophrenic patients. Tohoku Journal of Experimental Medicine 82: 124–39
18. Lhermitte J 1922 Syndrome de la calotte du pedoncule les troubles pysychosensoriels dans les lesions du mesencephale. Re ue Neurologique 38: 1359–65 November 9
19. Ojemann G, Mateer C. 1979 Human language cortex: Localization of memory, syntax, and sequential motor-phoneme identification systems. Science 205: 1401–3
20. Penfield W P, Perot P 1963 The brain’s record of auditory and visual experience. A final summary and discussion. Brain 86: 595–696
21. Pierce R C, Kalivas P W. 1997 A circuitry model of the expression of behavioral sensitization to amphetamine-like psychostimulants. Brain Res. Brain Res. Re . 25: 192–216
22. Porrino L J, Lucignani G, Dow-Edwards D, Sokoloff L 1984 Correlation of dose-dependent effects of acute amphetamine administration on behavior and local cerebral metabolism in rats. Brain Research 307: 311–20
23. Robinson T E, Kolb B 1999 Alterations in the morphology of dendrites and dendritic spines in the nucleus accumbens and prefrontal cortex following repeated treatment with amphetamine or cocaine. European Journal of Neuroscience 11: 1598–604
24. Scheich H, Walhausser-Franke E, Braun K 1991 Does synaptic selection explain auditory imprinting? In: Squire L R, Weinberger N M, Lynch G, McGaugh J L (eds.) Memory: Organization and Locus of Change. Oxford University Press, London, pp. 114–59
25. Silbersweig D A, Stern E, Frith C, Cahill C, Holmes A, Grootoonk S, Seaward J, McKenna P, Chua S E, Schnorr L,
26. Jones T, Frackowiak R S J 1995 A functional neuroanatomy of hallucinations in schizophrenia. Nature 378: 176–9
27. Spitzer M, Bohler P, Weisbrod M, Kischka U 1995 A neural network model of phantom limbs. Biological Cybernetics 72: 197–206
28. Wassermann E M, Wedegaertner F R, Ziemann U I, George M S, Chen R 1998 Crossed reduction of human motor cortex excitability by 1-Hz transcranial magnetic stimulation. Neuroscience Letters 250: 141–4
29. Wengel S P, Burke W J, Holemon D 1989 The sounds of silence? Journal of the American Geriatrics Society 37: 163–6
30. Wolf M E 1998 The role of excitatory amino acids in behavioral sensitization to psychomotor stimulants. Progress in Neurobiology 54: 679–70