01-anninos-

Acta neurol. belg., 2007, 107, 5-10
Original article
MEG evaluation of Parkinson’s diseased patients
after external magnetic stimulation
P. ANNINOS1, A. ADAMOPOULOS1, A. KOTINI1, N. TSAGAS2, D. TAMIOLAKIS3 and P. PRASSOPOULOS4 1Lab of Medical Physics, Medical School, Democritus University of Thrace, Alex/polis, Greece ; 2Lab of Nuclear Technology, Dept of Electrical Engineering and Computer Technology, Democritus Univ. of Thrace, Xanthi, Greece ; 3General Hospital of Chania, Crete, Greece ; 4Dept of Radiology, Medical School, Democritus University of Thrace, Alexandroupolis, Greece primate model of PD and, subsequently, by high- Magnetoencephalographic (MEG) recordings of frequency stimulation of the subthalamic nucleus in Parkinson’s diseased (PD) patients were obtained using PD patients, resulting in a remarkable reduction of a whole-head 122-channel magnetometer and analyzed symptoms (Krack et al., 2000 ; Volkmann et al., with Fourier statistical analysis. External transcranial magnetic stimulation (TMS) in the order of pico Tesla The subthalamic nucleus has a key-role in the was applied on the above patients with proper field pathophysiology of PD and is the primary target for characteristics (magnetic amplitude : 1-7.5 pT, frequen- high-frequency deep brain stimulation. The sub- cy : the a-rhythm of the patient : 8-13 Hz) which were thalamic nucleus rest electrical activity in PD, how- obtained prior to TMS. The MEG recordings after the ever, is still unclear. Priori et al. (2004), have test- application of TMS showed a rapid attenuation of the ed the hypothesis that pharmacological modulation high abnormal activity followed by an increase of the of subthalamic nucleus activity has rhythm specif- low frequency components toward the patients’ a-rhythm. The patients responded to the TMS with a feel- ic effects in the classical range of EEG frequencies, ing of relaxation and partial or complete disappearance below 50 Hz, and concluded that in the human sub- of tremor, muscular ache and levodopa induced dyskine- thalamic nucleus there are at least two rhythms sias as well as rapid reversed visuospatial impairment, below 50 Hz, that are separately modulated by which were followed by a corresponding improvement antiparkinsonian medication : one at low frequen- cies (2-7 Hz) and one in the beta range (20-30 Hz).
Key words : 122-channel magnetometer ; PD ; TMS ; Power changes elicited by antiparkinsonian med- ication in the alpha band were not significant. So,we have chosen alpha rhythm to apply externalmagnetic stimulation in our settings.
Introduction
The availability of MEG systems covering the whole scalp and methodological advances (Gross The current pathophysiological concept of et al., 2001) now allow investigation in more detail Parkinson’s disease (PD) postulates alterations of of the oscillatory network and mechanisms the interactions within the basal ganglia complex involved in PD tremor (Timmermann et al., 2003).
due to the loss of dopaminergic projections from Clinical applications of transcranial magnetic the substantia nigra to the striatum (Obeso et al., stimulation (TMS) were first reported by Baker et 2000). According to this model, pathological al. (1984) and has been widely used to assess pos- hyperactivity of the subthalamic nucleus drives the sible changes secondary to PD. The use of single- internal globus pallidus, which leads to an inhibi- and paired-pulse TMS, two varieties of the original tion of the ‘motor thalamus’ (ventro-lateral and technique, disclose multiple functional alterations ventro-anterior nuclei). Consequently, the output of of the corticospinal pathway (Cantello et al., 2002).
the thalamus to the sensorimotor cortex is reduced, The use of TMS in PD investigations began about resulting in hypokinesia. The involvement of other 10 years ago. Then it had become clear that TMS brain areas such as the supplementary and cingulate could provide information not only on the conduc- motor areas, premotor cortex, sensory cortices and tivity of corticospinal neurons, but also on other the cerebellum remains unclear in the described properties of the primary motor cortex, such as model. However, in the last years, this patho- excitability (Cantello et al., 1991). In turn, basic physiological concept of PD was corroborated by evidence strongly suggested that excitability was the successful treatment of a variety of PD symp- under the influence of multiple afferences to the toms by lesioning of the subthalamic nucleus in a motor cortex itself, among which those arising from the basal ganglia (Porter and Lemon, 1993).
Hence, a new insight arose into the pathophysiolo-gy of PD as well as of other movement disorders.
TMS has provided substantial new pathophysi-ological insights, which point to a central role ofthe primary motor cortex in the movement disordertypical of PD. Recently several clinical trials havesuggested the therapeutic efficacy of repetitiveTMS (rTMS) in patients with PD (Berardelli et al.,1999 ; Siebner et al., 2000 ; Strafella et al., 2001 ;Wassermann and Lisanby 2001 ; Khedr et al.,2003).
The goal of this study is to report the beneficial effects of external TMS (in the order of pico Tesla),on PD patients using MEG measurements and sta-tistical analytic techniques in the frequency FIG. 1. — The electronic device with the coils placed in two plastic plates by which the TMS is applied to PD patients.
Thirty PD patients (22 males, 8 females ; mean age 65 years, with range 49-80 years) were referred The time taken for each recording was 1 min.
to our laboratory by practicing neurologists from During the MEG recordings the subject was sitting January 2002 to December 2004 with symptoms of in a chair with his head covered by the helmet- akinesia, rigidity, or tremor and with EEG records shaped dewar. Four indicator coils attached to the before and after TMS. All the patients had been head of the individual subject determined the exact diagnosed to suffer from idiopathic PD on the basis position of the head with respect to the MEG of clinical observations and routine EEG record- sensors. The exact positions of the coils were deter- ings. The modified Hoehn and Yahr (H & Y) base- mined using a three dimensional digitizer. line status (Hoehn and Yahr, 1967), was stage 1.5 in Afterwards, external TMS in the order of pico 3 patients, stage 2 in 3 patients, stage 3 in Tesla was applied to PD patients with proper field 14 patients and stage 4 in 10 patients. The period characteristics (magnetic intensity : 1-7.5 pT ; fre- from diagnosis to the beginning of this study quency : the a-rhythm of the patient : 8-13 Hz), ranged from 1 to 3 years. None of them had a his- which were obtained prior to TMS using an elec- tory of other systemic neurological disease other tronic device (Anninos and Tsagas 1995 ; Anninos than PD, or implanted devices of pacemakers and et al., 1999 ; 2000 ; 2003). The coils of this device all had normal routine serum biochemical studies.
were placed on the patient’s scalp and weak mag- Patients had a neuroimaging study i.e CT (n = 12), netic fields, were applied for 6 minutes in total MRI (n = 7) or both (n = 3). In all cases written (2 minutes over each of the following areas : left informed consent for the methodology and the aim and right temporal regions, frontal and occipital of the study was obtained from all patients prior to regions, and over the vertex). This device consists the procedure. All patients were initially placed on of a generator to produce square waves of low fre- levodopa/carbidopa (Sinemet 25/250) (1 tablet quencies magnetic field in the range from 2-13 Hz twice daily), but due to progressive deterioration in to a group of coils of 1cm in diameter. The coils are their motor disability the dosage was increased to enclosed between two parallel plane surfaces in 3 1/2 tablets/day (1/2 tablet every 2 hours). All sub- such a way that their axis is situated perpendicular jects were off medication for 24 hours.
to these surfaces. The time between the first MEG and the MEG obtained after the application of TMS using a whole-head Neuromag 122 MEG system in a magnetically shielded room of low magnetic To confirm that the responses to TMS were noise with broadband (f > 10Hz) gradient noise reproducible, the patients were instructed to apply 5fT/(cmHz1/2) for the 95% of the channels and TMS with the same characteristics (2 times a week, max noise 10fT/(cmHz1/2), a broadband (1 Hz < f for 6 min total duration) nightly at home with the < 10 Hz) gradient noise 15fT/(cmHz1/2) for the electronic device (Fig. 1). In all patients placebo 95% of the channels and max noise 20fT(cmHz1/2) tests were also performed before the TMS and (Timmermann et al., 2003 ; Tonoike et al.,1998).
without energizing the device in order to evaluate The spontaneous MEG recordings were obtained the influence of the TMS. None of the patients from the PD patients using the 122-channel SQUID experienced any side effects during or after the pro- with sampling frequency of 256 Hz and filtered cedure. The statistical analysis of the results was with cut – off frequencies between 0.3 to 40 Hz.
obtained using the chi-square test and paired t-test.
Individual clinical data for each PD patient (N = 30). (A : abnormal ; P : partial normal ; N : normal diagnosis ; ance of very high amplitude power spectrum in thea-rhythm frequency). The difference was of statisti- Table I shows each patient’s clinical report and cal significance (p < 0.01, chi-square = 7.64). The their response to TMS. Based on an independent EEG and the MEG diagnosis before and after TMS chart review, they were divided into two groups is based on the appearance of a-rhythm amplitude according to the degree of their responsiveness to in their power spectra amplitude distribution TMS. The first group included patients who exhib- (tables I, II). Neuroimaging studies demonstrated ited only partial response (PR) to TMS (i.e., their diffuse cerebral atrophy in 5 patients and cortical tremor or muscular ache or dyskinesias recurred atrophy in one. Two patients exhibited small within 12 months after TMS and partial appearance ischemic infracts in the temporal lobes. No other of a-rhythm in their EEG denoted by low ampli- abnormalities were detected. After 1-2 months of tudes). The second group included patients who TMS, the H & Y stages showed significant decreas- demonstrated a favorable response (FR) to TMS es when compared with the baseline status. Three (i.e., they were free from the above symptoms for at patients showed no change of their H & Y status.
least one year after TMS and the appearance of a- They were stage 1.5 (2 of them) and stage 2 (the rhythm in their EEG denoted by high amplitudes).
remaining one). The scores of one patient stage 1.5, Using the above mentioned criteria table II was one patient stage 2, seven patients stage 3, and four formed. Twelve patients (40%) were classified as patients stage 4 decreased to averages 1.2 ± 0.3.
partial responders (PR) and the remaining 18 (60%) One patient stage 2 score decreased to 1.7. Five exhibited a favorable response (FR) to TMS patients stage 3, and three patients stage 4 scores (table II). Among the partial responders to TMS decreased to averages 1.6 ± 0.2. Two patients stage (41.67%), normal EEG (i.e., the appearance of high 3, and three patients stage 4 scores decreased to amplitude of power spectrum in the a-rhythm fre- averages 1.9 ± 0.1. The difference in the H & Y quency) was seen only in 5 patients, whereas 16 out status before and after TMS was of statistical of 18 patients who showed a favorable response to significance for the whole study group (p < 0.0001, TMS (88.88%) had normal EEG (i.e., the appear- paired t-test). Surprisingly, from the random choice same as the coherence found between motor cortexMEG, EEG or local field potentials and contralat- Classification of the examined PD patients according to their EEG and MEG diagnosis and their response to TMS eral electromyogram (EMG) during steady muscle contraction in humans and primates. (iii) The 15-30 Hz STN oscillations are diminished by volun- tary movement in a way analogous to the suppres-sion of human motor cortex beta EEG oscillations and motor cortex-muscle 15-30 Hz coherence inprimates and humans. (iv) Whilst we do not know if STN 15-30 Hz oscillations are present in non PDindividuals. Levy et al. (2002) show that treatmentwith apomorphine and levodopa suppresses theoscillations with a time course that correlates with of the patients only 4 of the 30 had abnormal EEG improvement of the “off” symptoms of PD.
BTMS, whereas the MEG BTMS was abnormal for (v) Suppression of 15-30 Hz STN oscillations with voluntary movement occurs independently ofchanges in the firing rate of STN neurons, indicat- Discussion
ing that their temporal pattern of discharge conveysadditional information to their firing rate. (vi) The The primary pathology of Parkinson’s disease 15-30 Hz oscillations are detected in the temporal (PD) is located in basal ganglia (DeLong, 1990).
patterning of STN neuron spike trains as well as at However TMS studies have demonstrated altered the level of local field potentials. (vii) The oscilla- excitability of the motor cortex in PD. Studies tions do not relate in any clear way to PD tremor using electrical and magnetic stimulation tech- and may relate more to mechanisms of akinesia.
niques have shown that the corticomotor neuron (Farmer, 2002) Coherence between STN area local connection is normal in PD (Dick et al., 1984).
field potentials and EEG is apparent in a wide This means that bradykinesia is not primarily the range of frequencies (theta : 3-7Hz, alpha : 8-13Hz, result of any deficit in the final output pathways of lower beta : 14-20Hz, and upper beta : 21-32Hz) the motor areas of the cortex. Most authors report- but activity in the alpha and upper beta bands ed that the motor cortex of patients with PD has the dominates (Fogelson et al., 2006) .
same threshold for stimulation as in healthy sub- Improvements, such as those that were found in jects (Ridding et al., 1995). However, when the the present study, are likely to be attributed to patients are tested at rest, the slope of the input-out- dopamine release, which is supported by an exper- put relationship between stimulus intensity and imental study in which repetitive TMS (rTMS) lead response size is steeper than normal. Perhaps as a to increased release of dopamine in the striatum result of this, voluntary contraction facilitates and frontal cortex (Ben-Shachar et al., 1997).
responses less than for normal subjects (Valls-Sole Strafella et al. (2001) showed that rTMS of the pre- et al., 1994). Although this could be the result of a frontal cortex induces the release of endogenous primary basal ganglia deficit, it seems probable that dopamine in the ipsilateral caudate nucleus as it could also be an attempt to compensate for the detected by positron emission tomography in slow recruitment of commands to move by making healthy human subjects. The rTMS-induced release it easier to recruit activity from a resting state of dopamine in the caudate nucleus could be a con- sequence of direct stimulation of the corticostriatal A work of Brown and colleagues (Brown et al., axons (Rothwell, 1997). GABA is the dominant 2001) has shown that in PD patients there is a inhibitory neurotransmitter of the motor cortex.
coherence between the motor cortex EEG and 15- Berardelli et al. (1999) recorded an increase in the 30 Hz subthalamic nucleus (STN) local field duration of the TMS-evoked SP during a 20-pulse potential oscillations. Thus the PD STN is driven train of suprathreshold rTMS in healthy volunteers by 15-30 Hz motor cortex oscillations. This leads as well as in PD patients. Mally and Stone (1999) to the hypothesis that the PD motor cortex-basal have reported sustained improvements in move- ganglia may be held abnormally in a 15-30 Hz ment-related measures with various regiments of oscillatory state ; yet these are the same coherent repeated TMS pulses administered with round coils frequencies as those detected between motor cortex over periods of weeks to months. Siebner et al. and muscle during postural maintenance in healthy (2000) recorded an increase in the duration of the humans. The study by Levy et al. (2002) contains a TMS-evoked SP in PD after 15 trains of 5-Hz number of important insights : (i) The frequency rTMS over the hand area. This means that 5-Hz range (15-30 Hz) of rhythms detected in STN is the rTMS is capable of inducing short-term change in same as that found in healthy humans to modulate the excitability of intracortical inhibitory circuitry motor unit activity during isometric muscle con- in PD patients. As dopamenergic drugs result in a traction. (ii) The 15-30 Hz frequency range is the similar modulation of the SP, the facilitatory effect of 5-Hz rTMS on intracortical inhibition might be BEN-SHACHAR D., BELMAKER R. H., GRISARU N., KLEIN E.
a candidate mechanism that mediates the beneficial TMS induces alterations in brain monoamines.
effect of 5-Hz rTMS of primary motor area in PD J. Neural Trans., 1997, 104 : 191-197.
BERARDELLI A., INGHILLERI M., GILLIO F., ROMEO S., In this study the patients’ responses to the TMS PEDACE F., CURRA A., MANFREDI M. Effects ofrepetitive cortical stimulation on the silent period were a feeling of relaxation and partial or complete evoked by magnetic stimulation. Exp. Brain Res., disappearance of muscular ache and levodopa- 1999, 125 : 82-86.
induced dyskinesias as well as rapid reversal of BERARDELLI A., ROTHWELL J. C., THOMPSON P. D., visuospatial impairment. This clinical improve- HALLETT M. Pathophysiology of bradykinesia in ment was followed by a corresponding improve- Parkinson’s disease. Brain, 2001, 124 : 2131-
ment and normalization of the MEG, recorded after the application of TMS. Assuming that the BROWN P., OLIVIERO A., MAZZONE P., INSOLA A., TONALI P., MEG of PD patients is a reflection of the DI LAZZARO V. Dopamine dependency of oscilla- pathogenesis in the substantia nigra, dopaminergic tions between subthalamic nucleus and pallidum functions and sympathetic ganglia, it appears that in Parkinson disease. J. Neurosci., 2001, 21 :
the application of the TMS has an immediate and beneficial effect on the dynamic condition of these ANTELLO R., GIANELLI M., BETTUCCI D., CIVARDI C., DE ANGELIS M. S., MUTANI R. Parkinson’s disease abnormally functioning neural structures (Sandyk rigidity : magnetic motor evoked potentials in a et al., 1991a ; 1991b ; 1991c ; 1991d ; 1992a ; small hand muscle. Neurology, 1991, 41 : 1449-56.
1992b ; 1992c ; 1992d ; 1992e ; 1992f ; 1992g ; CANTELLO R., TARLETTI R., CIVARDI C. Transcranial mag- 1992h). Although the striking beneficial effects of netic stimulation and Parkinson’s disease. Brain the application of the TMS on the clinical picture Res. Rev., 2002, 38 : 309-27.
of the PD patients are well observed, the mode of DELONG M. R. Primate models of movement disorders action of TMS in PD remains an open question.
of basal ganglia origin. Trends Neurosci., 1990, This question is difficult to be answered given the 13 : 281-285.
complexity of cellular, systemic and neuroen- DICK J. P., COWAN J. M., DAY B. L., BERARDELLI A., docrine effects of TMS on biological systems and KACHI T., ROTHWELL J. C., MARSDEN C. D.
Corticomotoneurone connection is normal in their potential impact on neurotransmitter func- Parkinson’s disease. Nature, 1984, 310 : 407-
tions. Despite all these and independent of their mechanisms of action, this method of magnetic FARMER S. Neural rhythms in Parkinson disease. Brain, stimulation may be considered an important non- 2002, 125 : 1175-1176.
invasive means in the management of idiopathic FOGELSON N., WILLIAMS D., TIJSSEN M., VAN BRUGGEN G., SPEELMAN H., BROWN P. Different functional loopsbetween cerebral cortex and the subthalamic area Acknowledgements
in Parkinson disease. Cerebral Cortex, 2006,
16 (1) : 64-75.
The authors would like to express their thanks and GROSS J., KUJALA J., HAMALAINEN M., TIMMERMANN L., appreciation to Dr. Carl Firley Vice President of IABC SCHNITZLER A., SALMELIN R. Dynamic imaging of (International Association of Biological Circuits) for his coherent sources : studying neural interactions in help and many stimulating discussions regarding this the human brain. Proc. Natl. Acad. Sci. USA, 2001, 98 : 694-9.
HOEHN M. M., YAHR M. D. Parkinsonism : onset, pro- gression, and mortality. Neurology, 1967, 17 :
KHEDR E. M., FARWEEZ H. M., ISLAM H., Therapeutic ANNINOS P. A., ADAMOPOULOS A., KOTINI A., TSAGAS N.
effect of repetitive transcranial magnetic stimula- Nonlinear Analysis of brain activity in magnetic tion on motor function in Parkinson’s disease influenced parkinson patients. Brain Topogr., patients. Eur. J. Neurol., 2003, 10 : 567-72.
2000, 13 : 135-44.
KRACK P., POEPPING M., WEINERT D., SCHRADER B., ANNINOS P. A., KOTINI A., ADAMOPOULOS A., TSAGAS N.
DEUSCHL G. Thalamic, pallidal, or subthalamic Magnetic stimulation can modulate seizures in surgery for Parkinson’s disease ? J. Neurol., 2000, epileptic patients. Brain Topogr., 2003, 16 : 54-64.
247 (Suppl 2) : 122-34.
ANNINOS P. A., TSAGAS N., JACOBSON J. I., KOTINI A. The LEVY R., ASHBY P., HUTCHINSON W. D., LANG A. E., biological effects of magnetic stimulation in LOZANO A. M., DOSTROVSKY J. O. Dependence of Epileptic patients. Panminerva Med., 1999, 41 :
subthalamic nucleus oscillations on movement and dopamine in Parkinson disease. Brain, 2002, ANNINOS P. A., TSAGAS N. Electronic apparatus for treat- 125 : 1196-1209.
ing epileptic individuals. US patent number MALLY J., STONE T. W. Improvement in Parkinsonian symptoms after repetitive transcranial magnetic BAKER A. T., JALINOUS R., FREESTON I. L. Non invasive stimulation. J. Neurol. Sci., 1999, 162 : 179-84.
magnetic stimulation of human motor cortex.
OBESO J. A., RODRIGUEZ-OROZ M. C., RODRIGUEZ M., Lancet, 1984, 1 : 1106-11.
MACIAS R., ALVAREZ L., GURIDI J., VITEK J., DELONG M. R. Pathophysiologic basis of surgery SANDYK R., ANNINOS P. A. Magnetic fields alter the for Parkinson’s disease. Neurology, 2000, 55 : S7-
circadian periodicity of seizures. Int. J. Neurosci., 1992f, 63 (3-4) : 265-74.
PORTER R., LEMON R. Corticospinal function and volun- SANDYK R., TSAGAS N., ANNINOS P. A., DERPAPAS K.
tary movement, Clarendon Press, Oxford, 1993, Magnetic fields mimic the behavioral effects of REM sleep deprivation in humans. Int. J. PRIORI A., FOFFANI G., PESENTI A., TAMMA F., Neurosci., 1992g, 65 (1-4) : 61-8.
BIANCHI A. M., PELLEGRINI M., LOCATELLI M., SANDYK R., TSAGAS N., ANNINOS P. A. Melatonin as a MOXON K. A., VILLANI R. M. Rhythm-specific proconvulsive hormone in humans. Int. J. pharmacological modulation of subthalamic Neurosci., 1992h, 63 (1-2) : 125-35.
activity in Parkinson Disease. Exp. Neurol., 2004, SIEBNER H. R., MENTSCHEL C., AUER C., LEHNER C., 189 : 369-379.
CONRAD B. Repetitive transcranial magnetic stim- RIDDING M. C., INZELBERG R., ROTHWELL J. C. Changes in ulation cause a short-term increase in the duration excitability of motor cortical circuitry in patients of the cortical silent period in-patients with with Parkinson’s disease. Ann. Neurol., 1995, 37 :
Parkinson’s disease. Neurosci. Lett., 2000, 284 :
ROTHWELL J. C. Techniques and mechanisms of action of STRAFELLA A. P., PAUS T., BARRETT J., DAGHER A.
transcranial magnetic stimulation of human Repetitive transcranial magnetic stimulation of cortex. J. Neurosci. Methods, 1997, 74 : 113-122.
the human prefrontal cortex induces dopamine SANDYK R., ANASTASIADIS P. G., ANNINOS P. A., TSAGAS N.
release in caudate nucleus. J. Neurosci., 2001, 1 ; Is postmenopausal osteoporosis related to pineal 21 (15) : RC157.
gland functions ? Int. J. Neurosci., 1992a, 62 (3-
TIMMERMANN L., GROSS J., DIRKS M., VOLKMANN J., FREUND H. J., SCHNITZLER A. The cerebral oscilla- SANDYK R., ANASTASIADIS P. G., ANNINOS P. A., TSAGAS N.
tory network of parkinsonian resting tremor.
Is the pineal gland involved in the pathogenesis of Brain, 2003, 126 : 199-212.
endometrial carcinoma. Int. J. Neurosci., 1992b, TONOIKE M., YAMAGUCHI M., KAETSU I., KIDA H., SEO R., 62 (1-2) : 89-96.
KOIZUKA I. Ipsilateral dominance of human SANDYK R., ANASTASIADIS P. G., ANNINOS P. A., TSAGAS N.
olfactory activated centers estimated from event- The pineal gland and spontaneous abortions : related magnetic fields measured by 122-channel implications for therapy with melatonin and whole head neuromagnetometer using odorant magnetic field. Int. J. Neurosci., 1992c, 62 (3-4) :
stimuli synchronized with respirations. Ann. N.Y. Acad. Sci., 1998, 855 : 579-590.
SANDYK R., ANNINOS P. A., TSAGAS N., DERPAPAS K.
VALLS-SOLE J., PASCUAL-LEONE A., BRASIL-NETO J. P., Pineal calcification and anticonvulsant respon- CAMMAROTA A., MCSHANE L., HALLETT M.
siveness to artificial magnetic stimulation in Abnormal facilitation of the response to transcra- epileptic patients. Int. J. Neurosci., 1991a, 60 (3-
nial magnetic stimulation in patients with Parkinson’s disease. Neurology, 1994, 44 : 735-
SANDYK R., ANNINOS P. A., TSAGAS N., DERPAPAS K.
Magnetic fields in the treatment of Parkinson’s VOLKMANN J., ALLERT N., VOGES J., WEISS P. H., disease. Int. J. Neurosci., 1992d, 63 (1-2) : 141-50.
FREUND H. J., STURM V. Safety and efficacy of pal- SANDYK R., ANNINOS P. A., TSAGAS N. Age-related dis- lidal or subthalamic nucleus stimulation in advan- ruption of circadian rhythms : possible relation- ced PD. Neurology, 2001, 56 : 548-51.
ship to memory impairment and implications for WASSERMANN E. M., LISANBY S. H. Therapeutic applica- therapy with magnetic fields. Int. J. Neurosci., tion of repetitive transcranial magnetic stimula- 1991b, 59 (4) : 259-62.
tion : a review. Clin. Neurophysiol., 2001, 112 :
SANDYK R., ANNINOS P. A., TSAGAS N. Magnetic fields and seasonality of affective illness : implications
for therapy. Int. J. Neurosci., 1991c, 58 (3-4) :
261-7.
ANDYK R., ANNINOS P. A., TSAGAS, N. Magnetic fields and the habenular complex. Int. J. Neurosci., 1991d, 59 (4) : 263-6.
SANDYK R., ANNINOS P. A. Attenuation of epilepsy with application of external magnetic fields : a case report. Int. J. Neurosci., 1992e, 66 (1-2) :

Source: http://mitosoft.eu/docs/referenz-therapie/anninos-2007.pdf