Pressure and Middle Cerebral Artery Blood Flow Velocity in Patients with Suspected Symptomatic Normal Pressure Hydrocephalus

Home , Middle cerebral artery

Jrournal of Neurology, Neurosurgery, and Psychiatry 1993;56:75-79 75 Simultaneous recording of cerebrospinal fluid J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.1.75 on 1 January 1993. Downloaded from pressure and middle cerebral artery blood flow velocity in patients with suspected symptomatic normal pressure hydrocephalus

D W Droste, J K Krauss

Abstract longer, have a higher amplitude, and appear CSF pressure (intracranial pressure, in plateau-like (therefore also referred to as "pla- one patient lumbar pressure) was mon- teau waves"). Lundberg also described itored continuously for one night in 23 C-waves particularly on the plateaux of the patients with suspected symptomatic nor- A-waves.5 B-waves are found with impaired mal pressure hydrocephalus (NPH) to CSF absorption in NPH and also less fre- identify patients who might benefit from quently in healthy persons,6 whereas A-waves subsequent shunt surgery. In 20 patients may indicate severely impaired CSF absorp- middle cerebral artery (MCA) blood flow tion.'- The origin of these waves is not yet well velocity by means of transcranial Doppler understood. Auer and Sayama observed rhyth- sonography (TCD) and CSF pressure mic pial vessel diameter changes correlating were recorded simultaneously. In three with B-waves.'0 Some authors claim a brain patients transcranial Doppler signals were stem centre is responsible for these oscilla- insufficient. Spontaneous changes in CSF tions." 2 Others stress the coincidence with pressure always paralleled changes in the respiratory and cardiovascular changes. 11-14 TCD signal. Equivalents of B-waves as Transcranial Doppler sonography permits well as intermediate waves (in between B- noninvasive measurement of the blood flow and A-waves), and C-waves could be iden- velocity in the large basal cerebral arteries.'5 tified easily and always appeared in phase. The diameter of these large arteries is assumed The Doppler signal, however, could not be to remain more or less constant.'6 Changes in used to evaluate the absolute changes in cerebral blood flow are caused by constriction CSF pressure. Fast Fourier Transform of or dilatation of the smaller resistance vessels. the Doppler signal was a useful tool to Therefore, changes of cerebral blood flow indicate the relative frequency of B-wave velocity in the large basal arteries correlate equivalents. In five patients the injection with changes of cerebral blood flow. ' In 5 of 10ml saline into the ventricle raised patients Maumer-Huppert et al showed that intracranial pressure considerably, but middle cerebral artery (MCA) blood flow hardly affected the MCA blood flow veloc- velocity and ICP were in phase during ity. Continuous TCD monitoring might be B-waves.6 Goh et al described a decrease in useful as a noninvasive screening proce- ICP with an increase in MCA blood flow http://jnnp.bmj.com/ dure in patients with suspected sympto- velocity and vice versa during A- and B-waves matic NPH before continuous invasive in three hydrocephalic children. In four other CSF pressure measurements are per- children the ICP and TCD signals were in formed. phase.'8 After subarachnoid haemorrhage, the signals took opposite courses. ' After severe (J Neurol Neurosurg Psychiatry 1993;56:75-79) head injury, they were in phase.20

Whether changes in cerebral blood flow/ on October 1, 2021 by guest. Protected copyright. volume are the cause or a consequence of A frequent problem encountered in neurology changes in ICP, there should be a relationship and neurosurgery is determining which between spontaneous ICP oscillations and the patients with suspected normal pressure TCD signal. The following study was designed University of hydrocephalus (NPH) will improve clinically to assess MCA blood flow velocity and CSF Freiburg, Freiburg, after shunt surgery. Positive criteria for the pressure simultaneously in patients with sus- Germany diagnosis of NPH include gait disturbance, pected NPH with special emphasis on the Department of of TCD of Neurology dementia and urinary incontinence, periven- characterisation signal equivalents D W Droste tricular lucency and large ventricles demon- spontaneous CSF pressure oscillations. Department of strated on CT, and clinical improvement after Neurosurgery lumbar puncture. Overnight recording of J K Krauss intracranial pressure (ICP) has become an Correspondence to: established method for selecting patients who Materials and methods Dr Droste, Department of Neurology, Atkinson will respond to shunt surgery. In our institutions patients with suspected Morley's Hospital, Copse Clinical improvement has been correlated NPH who might benefit from shunt surgery Hill, Wimbledon, London SW20 ONE, UK positively with recordings of A-waves or with are routinely submitted to an overnight record- Received 2 December 1991 frequent high amplitude B-wave activity.` ing of ICP. We examined 23 patients (6 men, and in revised form B-waves are spontaneous ICP oscillations with 17 women) with clinical signs of NPH. All 21 April 1992. Accepted 27 May 1992 a wave length of 0O5-2 minutes. A-waves last patients had ventricular enlargement ofvarious 76 Droste, Krauss

degrees on a cranial CT scan and all had at ICP was measured via an intraventricular

least one of the following symptoms: gait catheter of 2 mm diameter inserted into the J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.1.75 on 1 January 1993. Downloaded from disturbance, dementia, or urinary inconti- frontal horn of the left lateral ventricle through nence. Fourteen patients showed the complete a 2-mm burr hole, 2 cm parasagittal and 2 cm triad of symptoms. In two patients NPH was precoronal, drilled under local anaesthesia. thought to be secondary to meniu ngitis and The intraventricular catheter was secured to subarachnoid haemorrhage, respect:ively. Age the scalp with sutures. In one patient CSF ranged from 55 to 83 years, with the exception pressure was recorded via a lumbar catheter. of one patient who was 24 years old. The mean The transducer (Braun, Germany) was fixed age was 69 years. Informed corisent was over the bregma and connected to an amplifier obtained from the patients or their fiamilies, or (Servomed, Hellige, Germany) and a plotter both. In three elderly women we detected (Recomed, Hellige, Germany). There were no either no transcranial Doppler signalI or only a adverse events such as infection or bleeding. weak one which was not adequate for record- The foramina of Monro were considered to be ing. In the remaining 20 patients an overnight at Ocm CSF. The patients were in a supine (9-18 hours) recording of ICP (in oine patient position with the head flat. lumbar pressure) and MCA blood fli iw velocity For the TCD measurements we used the TC (9-12 hours, either right or left side) was 2000S transcranial Doppler (Eden Medizi- performed simultaneously. nische Elektronik GmbH, Uberlingen, Ger- many) and recorded only the envelope curve (that is, the maximal velocity), not the entire Doppler spectrum, numerically on the hard Fe -Ls. disc (sample rate 41 Hz). The highest middle cerebral artery blood flow velocity was sought at a depth of 45 to 55 mm through the ;,.' temporal window. The side measured was 100 - l 14t ;selected dependent on the quality of the signal and interference due to the catheter and 4.-t- sutures. We used the lowest ultrasound inten- TCD . '. 4t_ .._ sity still able to produce a satisfactory signal ' - Continuous measurement [cm/si '- 1I I' .I (<100 mW/cm2). . 4 -!-4 J i [CM/S'l,I.--1 -t was made possible by securing the probe in a I~ 1 _w_1 .1.I. . i, 1. head ribbon with adhesive tape. A person sat +{i|t;1__111,Ul'ttil}ljti'1. ,{ -1 V 30 -4t i*i |-ii4fh*4--tlfitl t| [r l l :l.. 1- .i beside the patient for the whole night, checked ,1*i.l;tt - 1' - -.--.-- !.1- -.L-.- I. -i-Xi- for any movement ofthe patient and refastened i, .1 I'. -. II, l-I ! the TCD probe when necessary. The CSF .Pl .-- - - ;-. - -.- -11".60 pressure was recorded throughout the night on I. -: __:. .11. '..I.1I, -". a plotter at a speed of 5 mm/min and inter- 0 .11, L1~ ~' ~ mittently at a higher speed (1 mm/s, 10 mm/s). I ICP t [s] The transcranial Doppler signal was recorded 0 [cm CSF] 5 10 on the hard disc of the TC 2000S and in 15 patients also on the same sheet of paper as the Figure I Simultaneous recording of middle cerebral artery blood flow velocity by CSF pressure using the second channel of the transcranial Doppler (TCD) and intracranial pressure (ICP). Patient 14. plotter. http://jnnp.bmj.com/ The waveforms are not totally parallel. A rhythm of about 3 cardiac cycles of rwavelength is visible in the ICP recording, but not in the TCD recording. There is a sligh,tdelay of In five patients Doppler signals and CSF about O ls between the ICP and TCD signals during one cardiac cycle. pressure were recorded simultaneously during a bolus injection of 10 ml saline via the ventricle catheter. Sixteen patients had ventriculoatrial or ventriculoperitoneal shunting two to four

weeks after the CSF pressure / TCD record- on October 1, 2021 by guest. Protected copyright. ing.

Results + In all patients the initial baseline intracranial / lumbar pressure was less than 20 cm CSF, : which was consistent with a diagnosis of "Anormal" pressure hydrocephalus. The Dop- pler recording was more susceptible to movement artefacts than the pressure registration. However, the artefacts could be reliably identified. In the 15 parallel recordings the course of the Doppler signal was always almost parallel to and in phase with the course of the CSF ICP t I in I pressure. c_rni CS F 1 2 At a speed of 10 mm/s during one cardiac cycle, there was only a delay of about 01 s Figure 2 Patient 18. Simultaneous recording of MCA blood flow velocity by TCD and between the systolic peak and the dicrotic ICP Permanent B-wave activity. Increase and decrease in the two signals are7 in phase. incisura of the CSF pressure and of the blood Simultaneous recording of CFP and MCA blood flow velocity in patients with suspected SNPH 77

flow velocity (fig 1). The waveform was ...... slightly ....;... -;.... :

...... - J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.1.75 on 1 January 1993. Downloaded from . _...... - different. A rhythm with a wavelength of about three heartbeats was seen in the pressure recording, but not in the Doppler recording ..... (probably corresponding to respiration). * .. The two recordings were also almost parallel at a of a ''''a'----- speed 1 mm/s. This speed provides -...... - , good image of the B-waves and their TCD equivalents in detail (fig 2). The routinely used T------..... speed of 5 mm/min is appropriate for visual- + ising intermediate waves (see below) and long ... + -t-- runs of B-waves and their TCD equivalents 30 ...... ::::: -'.. : ; - (figures 3a, b, c). In all of the 20 patients we observed B-waves with various wavelengths and amplitudes hav- _T ing sine-typed or ramp-typed shape. The rela- 0 - tive frequency was variable. In four patients ICP t [min] there were a few waves lasting longer than 2 0 Icm CSFI 8 minutes and 40-60 cm CSF high. These latter waves (considered as intermediate waves, between A- and B-waves) were clearly dis- tinguishable from the surrounding smaller B-waves, however, they never reached a pla- O .. teau of several minutes as the A-waves des- 46b -L i-,LI I i i-m &sA isiss cribed by Lundberg.5 Some of them looked like large ramp-typed B-waves, others resembled A-waves with only a short plateau. C-waves were observed in only two patients 40 i V TTVU (fig 3c). To confirm the visual impression of an WIV almost parallel course of CSF pressure and TCD MCA blood flow velocity, we performed meas- [cm/s] urements at this speed (5mm/min) in each of the 15 patients (ICP and TCD on one sheet of , paper) of the systolic value of the peak and the 30 -i iA :. jsii diastolic value of the following dip in 5 TCD B-wave equivalents (and not knowing that there also is a B-wave in the pressure recording). Then we measured the corresponding 0 values of the pressure recording. We did the r-- ICP t [min] same for 5 other B-waves seen first on the pressure recording and then measured the 0 [cm CSF] 20 corresponding Doppler values. In all of these

150 B-waves/B-wave equivalents maximum http://jnnp.bmj.com/ and minimum values were reached in both recordings at the same time. There was no

......

...... , . ., + w ...... B-wave 0 recording that was not reflected by a 1 t '- '' t ... -- B-wave equivalent in the TCD and vice versa. We also measured 2 x 11 intermediate waves and theirTCD equivalents in four patients and 100 - observed the same result. Intermediate waves and their TCD equivalents were higher than on October 1, 2021 by guest. Protected copyright. B-waves and their TCD equivalents in the TCD same patient. TCD artefacts due to movement Icm'sl with probe displacement were reliably detected. In two restless patients who produced frequent artefacts, some of these artefacts (for example, due to yawning, micturition, or ...... A 0-. . . .. --. - ...... A _Aw A , coughing) resembled spontaneous ICP oscillations (probably due to pressing and sub- *:._.tr:u--- * - *; | * ->_ 1|-- *--: t st S...... ;1. | t MIM", riqt.....F '*.' ...ir''1-W-IFWT---.;_-* I sequent rise of intrathoracic and venous pressure) and were not correlated with changes .... O I -- --- t [mi.n] in MCA blood flow velocity. The occurrence of ICP ml' these artefacts was noted by the night watch. 0 [cm CSF] 20 There was no difference between the lumbar

Figure 3 Simultaneous recording of MCA blood flow velocity by TCD and I( -P in three and ventricular pressure measurements for different patients. spontaneous CSF pressure oscillations and 3a (top): Permanent B-wave activity. Speed was changed from 1mmls to Smi ,z/min and their TCD equivalents. back to Immls. Same patient as in fig. 2 (patient 18). In each of 3b (middle): Patient 4. Permanent B-wave activity. the four patients with inter- 3c (bottom): Patient 14. Several B-waves and an intermediate wave. C-waves on top of mediate waves we made one graph with 10 the intermediate wave. B-waves and 1-10 intermediate waves sampled 78 Droste, Krauss

randomly from the recording. The MCA blood 0 flow velocity constituted the one coordinate, J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.1.75 on 1 January 1993. Downloaded from the CSF pressure the other. Each wave con- sisted of two points, the peak (systolic value) and the following dip (diastolic value), as these 50 parameters were easily measurable in the recording. Diastolic value of the peak and systolic value of the dip were frequently super- imposed by surrounding values. We chose these four patients, as they offered a wider range of velocity and pressure due to the intermediate waves. In three patients there was 30 a good intraindividual correlation of CSF pressure andTCD signal within a certain range (0-30 cm CSF). Above 30 cm CSF MCA blood flow velocity rose more slowly. In the 0 X, 1 Oml fourth patient there was also a linear correlation above 30 cm of CSF. There was no ICP t [min] interindividual correlation between changes in ° [cm CSF] 10 MCA blood flow velocity and changes in CSF pressure. One patient might show, for example, Figure 5 Simultaneous recording of MCA blood flow velocity by TCD and ICP Patient 17. Injection of IOml B-wave equivalent activity of high amplitude in saline into the lateral ventricle (arrow) raised the ICP the TCD recording, corresponding to B-waves considerably, but hardly affected the TCD signal. of low amplitude, and another patient the opposite. With the TC 2000S it is possible to create Discussion ASCII "export" files off line that include the Equivalents of spontaneous CSF pressure values of the envelope curve (41 values per oscillations (B-waves, intermediate waves and second) or the mean of the values next to each C-waves) in patients with suspected NPH can other (data compression). In 16 patients we be detected by continuous transcranial Dop- performed a Fast Fourier Transform of 6 pler recordings of the middle cerebral artery. continuous early morning hours of velocity Conversely, forms appearing as pressure waves recording. The data were compressed to a on a TCD recording correspond to similar sample rate of 41 /min, which was sufficient to waves on the pressure recording. With a person detect B-waves with a wavelength of 0 5-2s. We sitting beside the patient artefacts can be show a representative power spectrum of a reliably identified. Therefore, TCD might be patient with almost permanent B-wave activity useful as a noninvasive screening procedure for (fig 4). patients with suspected NPH, as it provides The other spectra either resembled that information on the relative frequency and shown in fig 4 or had no or only a smaller peak shape of spontaneous CSF pressure oscilla- between 0 5 and 2 minutes of wavelength. tions. Invasive CSF pressure measurement is The bolus injection of 10 ml saline into the mandatory to confirm a positive result and to at lateral ventricle of five patients hardly affected assess the absolute level of CSF pressure http://jnnp.bmj.com/ the Doppler signal. It did, however, elevate the baseline and during oscillations. The relative ICP considerably (fig 5). frequency of B-waves, in particular, is con- spectral sidered to be a predictor for shunt responsive- power ness.' 4 In patients with suspected NPH who show only few B-wave equivalents during continuous TCD recording, CSF pressure recordings may possibly give no further information concerning the relative frequency of on October 1, 2021 by guest. Protected copyright. B-waves. On the other hand, we propose that those patients with frequent activity of B-wave equivalents or equivalents of intermediate waves should have further investigation with CSF pressure monitoring. Intermediate and C-waves were rare in our recordings. Any existing intermediate waves were relatively small in wavelength and did not reach a long plateau as the A-waves described by Lundberg. Lundberg reported A- and C-waves to be present in patients with a considerable increase in the mean CSF pressure (for example, due to space-occupying lesions).5 All of our patients had "normal" pressure hydrocephalus. Three of our patients initially showed a wavelength [min] linear correlation between TCD signal and CSF pressure, however, at levels above approx- Figure 4 Fast Fourier Transform of the transcranial Doppler recording over 6 hours. A 30 cm CSF with increasing pressure peak of the spectral power at 40-50s corresponding to almost permanent B-wave activity imately In this patient (patient 3). the concomitant increase in blood flow velocity Simultaneous recording of CFP and MCA blood flow velocity in patients with suspected SNPH 79

was less pronounced. In another patient there 1 Symon L, Dorsch NWC. Use of long-term intracranial pressure measurement to assess hydrocephalic patients J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.1.75 on 1 January 1993. Downloaded from seemed to be a linear correlation above 30 cm prior to shunt surgery. J Neurosurg 1975;42:258-73. 2 Pickard JD. Adult communicating hydrocephalus. BrJ Hosp CSF, too. Med 1982;27:35-44. There was an almost parallel increase and 3 Black PM, Ojemann RG, Tzouras A. CSF shunts for decrease in the two dementia, incontinence, and gait disturbance. Clin Neuro- signals during spontaneous surg 1985;32:632-5 1. oscillations. This is consistent with the findings 4 Graff-Radford NR, Godersky JC, Jones MP. Variables et Goh predicting outcome in symptomatic hydrocephalus in the reported by Mautner-Huppert al,6 et elderly. Neurology 1989;39:1601-4. al 8 and Newell et al.2" The absence of marked 5 Lundberg N. Continuous recording and control of ven- blood flow tricular fluid pressure in neurosurgical practice. Acta MCA velocity changes during Psychiatr Neurol Scand 1960;Suppl 149:1-193. injection of normal saline into the lateral 6 Mautner D, Dimagl U, Haberl R, Schmiedeck P, Garner C, ventricle with a concomitant increase in CSF Villringer A, Einhaupl KM. B-waves in healthy persons. In: HoffJT, Betz AL, eds. Intracranial pressure VII. Berlin, pressure supports the hypothesis that sponta- Heidelberg: Springer-Verlag, 1989:209-12. neous CSF oscillations are of vaso- 7 Hayashi M, Kobayashi H, Kawano H, Handa Y, Yamamoto pressure S, Kitano T. ICP patterns and isotope cisternography in genic origin. The origin of vasodilatation at the patients with communicating hydrocephalus following a rupture of intracranial aneurysm. J Neurosurg 1985;62: beginning of B-wave and of vasoconstriction 220-6. at the end of a B-wave as described by Auer 8 Yokota A, Matsuoka S, Ishikawa T, Kohshi H, Kajiwara H. and and their relation or Overnight recording of intracranial pressure and electro- Sayama9 (cause encephalography in neurosurgical patients. Part I: Intra- effect) to raised intracranial pressure remain cranial pressure waves and their clinical correlations unclear. There are different on the (abstract). Sangyo Ika Daigaku Zasshi 1989;11:371-81. opinions 9 Hayashi M, Handa Y, Kobayashi H, Kawano H, Ishii H, role of rhythmic hyperventilation and hypo- Hirose S. Plateau waves and CSF circulation. The eighth ventilation in international symposium on intracranialpressure (Abst 202). (and subsequent changes pCO2 Rotterdam. 16-20 June 1991. Abstract 202. and vessel diameter, for example, in Cheyne- 10 Auer LM, Sayama I. Intracranial pressure oscillations Stokes vasodilatation (B-waves) caused by oscillations in cerebrovascular vol- respiration), neurogenic ume. Acta Neurochir 1983;68:93-100. and vasoconstriction, and changes in blood 11 Higashi S, Yamamoto S, Hashimoto M, Fujii H, Ito H, and heart rate. and Kogure Y, Tokuda K. The role of vasomotor center and pressure Lundberg4 Maeda adrenergic pathway in B-waves. In: HoffJT, Betz AL, eds. et al12 reported that A-waves appear not only Intracranial pressure VII. Berlin, Heidelberg: Springer- but also artificial Verlag, 1989:220-4. during spontaneous, during 12 KogureY, Fujii H, Higgashi S, Hashimoto M,Tokuda K, Ito respiration, but that B-waves are often effaced H,Yamamoto S. Electroencephalographic changes during with the start of artificial ventilation. Shima et plateau waves induced by electrical stimulation of the brain stem. In: HoffJT, Betz AL, eds. Intracranial pressure al22 described hypoventilation during plateau VII. Berlin, Heidelberg: Springer-Verlag, 1989:235-7. waves in humans. in heart rate were 13 Einhaupl KM, Garner C, Dirnagl U, Schmiedeck P, Kufner Changes G, Rieder J. Oscillations of ICP related to cardiovascular minimal. Newell et al20 found B-waves in parameters. In: Miller JD, Teasdale GM, Rowan JO, ventilated Hashimoto et Galbraith S, Mendelow AD eds. Intracranial pressure VI. artificially patients. Berlin, Heidelberg: Springer-Verlag, 1986:290-7. al"4 described apnoea at the beginning of a 14 Hashimoto M, Higashi S, KogureY, Fujii H, Tokuda K, Ito followed and H, Yamamoto S. Respiratory and cardiovascular oscilla- B-wave, by respiration increasing tions during B-waves. In: Hoff JT, Betz AL, eds. heart rate and blood pressure: the peak of a Intracranial pressure VII. Berlin, Heidelb-erg: Springer- B-wave was reached before the of heart Verlag, 1989:217-9. peak 15 Aaslid R, MarkwalderTH, Nornes H. Noninvasive trans- rate and blood pressure. Borgesen and cranial Doppler ultrasound recording of flow velocity in found no correlation between arte- basal cerebral arteries. J Neurosurg 1982;57:769-74. Espersen2' 16 Huber P, Handa J. Effect of contrast material, hypercapnia, rial pCO, and A- and B-waves, excluding hyperventilation, hypertonic glucose and papaverine on in to be a causative the diameter of the cerebral arteries. Invest Radiol changes respiration factor. 1967;2: 17-32. Einhupl et al " reported hyperventilation to 17 Bishop CCR, Powell S, Rutt D, Browse NL. Transcranial Doppler measurement of middle cerebral blood flow

coincide with the B-wave in with http://jnnp.bmj.com/ peak patients velocity. A validation study. Stroke 1986;17:913-5. and without Cheyne-Stokes respiration. Ani- 18 Goh D, Minns RA, Pye S, Steers J. Continuous intra- mal conducted Maeda et ventricular pressure and simultaneous cerebral blood flow experiments by al2l velocity (CBFV) measurements during sleep in hydro- and Kogure et al 12 using electrical stimulation cephalic children. The eighth international symposium on showed evidence of a brainstem centre intracranial pressure (Abst 221). Rotterdam, 16-20 June being 1991. responsible for neurogenic vasodilatation and 19 Hashimoto T, Nakamura N, Kanki T, Abe S. ICP and flow vasoconstriction. These are dynamics after subarachnoid haemorrhage. The eighth findings, however, international symposium on intracranialpressure (Abst 109). difficult to interpret with regard to the cause of Rotterdam, 16-20 June 1991. as electrical 20 Newell DW, Stooss R, Aaslid R, Reulen HJ. Spontaneous on October 1, 2021 by guest. Protected copyright. spontaneous ICP oscillations, fluctuations in cerebral blood flow as a cause of B-waves. stimulation is non-physiological. The fact that The eighth international symposium on intracranial pressure ICP rises in II and in REM (Abst 127). Rotterdam, 16-20 June 1991. sleep stage sleep 21 Maeda M, Matsuura S, Tanaka K, Katsuyama J, Nakamura and the fact that spontaneous ICP oscillations H, Sakamoto H, Nishimura S. Effects of electrical are more in REM stimulation on intracranial pressure and systemic arterial frequent sleep'82' sug- blood pressure in cats. Part I: Stimulation of brain stem. gests that spontaneous CSF pressure oscilla- Neurol Res 1988;10:87-92. tions are of cerebral mediated 22 Shima K, Ueno K, Chigaski H, Ishii S. Comparative study origin by with experimental and clinical plateau waves relating to changes in vessel diameter. sleep. In: Hoff JT, Betz AL, eds. Intracranial pressure VII. The Fast FourierTransform of aTCD Berlin, Heidelberg: Springer-Verlag, 1989:241-4. signal 23 Borgesen SE, Espersen JO. The correlation between A- and over 6 hours is helpful, since the frequency and B-waves and arterial CO2 tension. In: Miller JD, Teasdale of B-wave are GM, Rowan JO, Galbraith S, Mendelow AD, eds. amplitude equivalents taken into Intracranial pressure VI. Berlin, Heidelberg: Springer- account. As shunting in NPH results in a Verlag, 1986:298-304. reduction of the and the relative 24 Ogashiwa M, Takeuchi K. Intracranial pressure changes amplitude during sleep in man. No To Shinkei 1983;35:123-9. frequency of B-waves,26 TCD and Fast Fourier 25 Yokota A, Matsuoka S, Ishikawa T, Kohshi K, Kajiwara H. Transform also be useful in Overnight recordings of intracranial pressure and electro- might controlling encephalography in neurosurgical patients. Part II: Chan- the result of shunting. ges in intracranial pressure during sleep. Sangyo Ika Daigaku Zasshi 1989;11:383-91. We are very grateful to Eden Medizinische Elektronik GmbH, 26 Tanaka K, Nishimura S. The importance of outflow Uberlingen, Germany for supplying us with the TC 2000S and resistance of the shunt system for elimination of B-waves. to Professor Schulte-Monting for the biomathematical inter- In: Hoff JT, Betz AL, eds. Intracranial pressure VII. Berlin, pretation of the data. Heidelberg: Springer-Verlag, 1989:368-73.