Original Research JOURNAL OF MAGNETIC RESONANCE IMAGING 30: (2009)

Transcription

1 JOURNAL OF MAGNETIC RESONANCE IMAGING 30: (2009) Original Research Value of Quantitative MRI Biomarkers (Evans Index, Aqueductal Flow Rate, and Apparent Diffusion Coefficient) in Idiopathic Normal Pressure Hydrocephalus Samuel E.S. Ng, MBBS, FRCR, 1 * Angela M.S. Low, DCR (R), 1 Kok Kee Tang, MBBS, FRCS, 2 Y.H. Chan, PhD, 3 and Robert K. Kwok, MBBS, FRCR 1 Purpose: To define the value of Evans index (EI), aqueductal flow rate (FR), and apparent diffusion coefficient (ADC) in the diagnosis of normal pressure hydrocephalus (NPH) and to assess the ability of these markers preoperatively to predict shunt response. To shed some light as to the mechanisms responsible for the symptoms of NPH. Materials and Methods: Preoperative EI, FR, and ADC readings in nine cases of clinically diagnosed NPH were compared with those of age- and gender-matched controls. Similar pre- and postoperative readings of responders and nonresponders were subsequently compared. Results: Compared with the controls, all measurements were statistically significant except for peak systolic flow rate (psfr), which was near statistical significance. Comparison of pre- and postoperative readings of responders and nonresponders revealed a decrease in ADC in all responders (P 0.032). Subdural hemorrhage was found in all nonresponders (P 0.012). Conclusion: For patients presenting with signs and symptoms of NPH, readings on MRI greater than 0.3, 10 ml/ min, 9.0 ml/min, and mm 2 /s for EI, peak diastolic flow rate (pdfr), psfr, and ADC, respectively, add further weight to the diagnosis. The strong correlation between shunt response and ADC decline support our hypothesis that water accumulation in the cerebrum is the major cause for the symptoms of NPH. The presence of subdural hemorrhage in all nonresponders raises suspicion of decreased compliance as the other major cause. Key Words: idiopathic normal pressure hydrocephalus; ventriculo-peritoneal shunt; peak systolic flow rate; Evans index; apparent diffusion coefficient; subdural hematoma J. Magn. Reson. Imaging 2009;30: Wiley-Liss, Inc. 1 Department of Radiology, Mount Elizabeth Hospital, Singapore. 2 K.K. Tang Adult and Paediatric Neurosurgery, Singapore. 3 Biostatistics Unit, Yong Loo Lin School of Medicine, National University of Singapore, Singapore. Contract grant sponsor: Biomedical Research Council/ A*STAR, Singapore; Contract grant number: 01/1/27/19/200. *Address reprint requests to: S.E.S.N., Department of Radiology, Mount Elizabeth Hospital, Singapore samnes@gmail.com Received November 2, 2008; Accepted June 5, DOI /jmri Published online in Wiley InterScience ( IDIOPATHIC NORMAL PRESSURE hydrocephalus (NPH) is a debilitating condition that afflicts predominantly the elderly (1,2). The patient typically presents with the following triad: (i) loss of bladder and bowel control (wet); (ii) abnormal broad based gait (wobbly); and (iii) loss of memory and mental function (wonky). The diagnosis is supported by the radiologic finding of dilated ventricles (defined by an elevated bifrontal or Evans index, EI) (3) without a demonstrable cause. It commands the attention of the medical community because NPH-related dementia is possibly the only surgically reversible dementia. Patients with newly diagnosed NPH typically respond to ventriculo-peritoneal shunting (VPS) (4). How shunting reverses the dementia and other symptoms of hydrocephalus are not known. As such, an understanding of the former is critical because this can shed light on how other forms of dementia can be reversed. Conventional diagnostic investigations have relied on computed tomography (CT) derived EI, intracranial pressure monitoring and cerebrospinal fluid (CSF) infusion tests (5 9). The latter two tests are invasive and costly. Their reliability and reproducibility are limited (2). MRI as an investigative tool has been gaining popularity, initially for its noninvasive, nonionizing approach. It has also evolved from qualitative assessment to include quantitative evaluation, making it ideal for interrogation of both structural and functional changes in NPH. This study focuses on the role of three biomarkers (EI, aqueductal flow rate [FR], and apparent diffusion coefficient [ADC]) in assessing NPH. All three are easily acquired or computed using sequences and software that are widely available in most superconducting MRI scanners. Two of these, EI and FR, have been used to evaluate different aspects of NPH. EI is a reflection of ventriculomegaly and is a critical diagnostic marker of NPH (7,8). FR, on the other hand, has been reported as a marker for shunt response. An elevated CSF FR is associated with a positive response to shunting (10,11) Wiley-Liss, Inc. 708

2 MRI Biomarkers in Idiopathic NPH 709 The role of ADC in NPH has only been recently highlighted (12) as a reading to understand the evolution of the disease. Accumulation of fluid intracranially (be it intra- or extracellular) can be easily quantified with ADC mapping, from the diffusion weighted imaging (DWI) sequence. The objectives of this study are three-fold. The first seeks to define the value of EI, FR, and ADC in the diagnosis of NPH. This is done by comparing the measurements of clinically diagnosed NPH with those of age- and gender-matched controls. The second is to assess the ability of these markers preoperatively to predict shunt response. The third objective seeks to shed some light as to the mechanisms responsible for the symptoms of NPH. NPH is presumably related to increased resistance to CSF outflow and/or reduced CSF absorption through the arachnoid villi. Blockage can also occur at the level of the venous sinuses and/or the subarachnoid spaces. Whatever the cause, ventriculomegaly, hyperdynamic CSF flow, and extracellular CSF accumulation in the intracranial tissue can be seen as consequences to this blockage. We believe that the accumulation of CSF within the intracranial tissue is a major determinant of the symptoms of NPH, more so than ventriculomegaly or hyperdynamic flow. We seek to test our hypothesis by comparing clinical response to pre- and postoperative measurements of the three most relevant and easily measurable biomarkers related to NPH, i.e., the EI, FR, and ADC. If the former (i.e., CSF accumulation as a major determinant) were true, all responders would post an interval postshunt decrease in ADC, with the reverse holding true for nonresponders. If the latter (i.e., ventriculomegaly or hyperdynamic flow as the major determinant) were true, all responders should post an interval decrease in FR and EI. Moreover, the reverse would hold true for nonresponders. MATERIALS AND METHODS Nine patients (age, years; mean, 70.2 years; male:female, 2:1) with a clinical diagnosis of idiopathic NPH were studied and compared with nine age- and gender-matched controls. The study was approved by the local ethics committee. All patients and/or relatives gave informed consent and all procedures were in accordance with institutional guidelines. All scans were performed with a 1.5 Tesla (T) wholebody MR scanner (GE Signa, General Electric Medical Systems, Milwaukee, WI) equipped with high-performance gradients by using a manufacturer-supplied quadrature head coil. Scans were acquired between 1 and 7 days before VPS, as well as 6 to 12 months after VPS. Aside from conventional orthogonal sections, the following MR measurements were made. EI Maximal frontal horn ventricular width divided by the transverse inner diameter of the skull. Ventriculomegaly is defined as a reading of 0.3 or greater (3). FR In our choice of hydrodynamic metric, we considered velocity (10,13 15), flow (10,14 17), and aqueductal stroke volume (14,18) parameters. At the time of this study, only flow velocity and flow rate were available in the analysis package. Both metrics are simple and fast to use, as well as reliable (10,15,17). Flow rate was preferred over flow velocity, because the introduction of the aqueductal area in FR computation corrects for over- or undertilting of the axial slice placement over the aqueduct (17). The initial acquisition was made at the level of the cerebral aqueduct using prospective cardiac gating phase contrast MRI. The level of interest was centered to the inferior colliculus, perpendicular to a line drawn through the distal aqueduct on the sagittal scout (Fig. 1). Parameters included: repetition time/echo time (TR/TE), 19/8 ms; FA 20 ; section thickness 7mm; matrix matrix; field of view (FOV) 16 cm; Venc 20 cm/s; flow compensation and 16 cardiac phases by means of peripheral gating with finger plethysmography. The data were then processed by means of a flow analysis program (General Electric, Milwaukee, WI), which generated an average flow rate in ml/min for each cardiac phase. The aqueductal area of measurement ranged from 6.88 to 9.57 mm 2. The pixel range was 22 to 31. A curve of these readings was generated and the peak diastolic flow rate (pdfr) and peak systolic flow rate (psfr) read off this curve (10,17). Flow rates were deemed normal or hyperdynamic based on computations from a previous study (Fig. 2a,b) (19). ADC Value of the Periventricular Region A DWI sequence was applied over the entire brain TR/ TE 10/0.125 s, FOV 34 cm, matrix , slice thickness 5 mm, b 1000 s/mm 2. ADC was computed by placing a 45 mm 2 circular region of interest (ROI) in the periventricular region adjacent to the tip of the frontal and occipital horns of each lateral ventricle (Fig. 3). The average of these four readings was then computed. Based on prior studies (20 22), ADC was considered elevated if the reading was mm 2 /s or greater. Similar readings of 9 age and gender matched controls were also recorded. All patients were shunted using a medium-pressure (Medtronic PS Medical, Goleta, CA), CSF contoured valve system. Clinical Grading of Symptoms Pre- and postshunting The three primary symptoms of NPH (i.e., gait disturbance, urinary incontinence, and cognitive impairment) were assessed by two doctors using the scale developed by Krauss et al (22). Score assignment was based on the clinical examination as well as interviews with the patient and the primary caregiver. Based on this scale, patients were ascribed a pre- and postshunt score.

3 710 Ng et al. Figure 1. Sagittal scout image, on which the level of the phase contrast cine series is determined. The level is at the inferior colliculus, along a plane perpendicular to the long axis of the cerebral aqueduct. Figure 2. a: Quantitative analysis demonstrating normal flow in a healthy volunteer. b: Quantitative analysis demonstrating marked hyperdynamic flow rates in a normal pressure hydrocephalus patient.

4 MRI Biomarkers in Idiopathic NPH 711 Figure 3. Apparent diffusion coefficient map of a patient with normal pressure hydrocephalus. A total improvement index (TII) was also computed, reflecting the ratio of postoperative improvement over the severity of the symptoms. Patients with a positive TII (i.e., 0) were deemed responders. Those with a score of 0 were deemed nonresponders. Significant postshunt findings (Other findings [Oth Fi], e.g., infarcts, subdural hemorrhage (SDH), and so on) were also recorded. Statistical Analysis All statistical analyses were performed using SPSS 15.0 with statistical significance set at P Wilcoxon Signed Rank test was performed to compare the differences in shunt measurements between patients and their matched controls. Fisher s Exact test was used to assess the differences between responders and nonresponders. Mann Whitney U test was applied to compare the differences in the % change in shunt measurements between the two groups and adjusted for age and gender using multiple regression analysis. RESULTS Comparing Preshunt Measurements Between NPH Cases and Controls The demographic data and the preshunt MR measurements (EI, pdfr, psfr, and ADC) of NPH cases and their (age- and gender-matched) controls are summarized in Table 1. The statistical comparison of these measurements between patients and controls are summarized in Table 2. Compared with the controls, all measurements were statistically significant except for psfr, which was near statistical significance. Comparing Pre- and Postshunt Measurements Between Shunt Responders and Nonresponders Patients 1 6 posted a positive total improvement index and were deemed shunt responders. Patients 7 9 posted a zero or negative total improvement index and were deemed shunt nonresponders. The demographic data, pre- and postshunt MR measurements (EI, pdfr, psfr, and ADC) of NPH cases for shunt responders and shunt nonresponders are summarized in Tables 3 and 4, respectively. All six responders showed interval decrease in ADC compared with all the nonresponders posting an elevation. For the responders, the percentage decrease ranged from 0.1% to 12.3%. This drop was low for cases 1 and 2, where the difference was less than 6.7%. For the other four responders, the difference was equal or greater than 6.7%. The percentage elevation for the nonresponders ranged from 3% to 10%. The other two

5 712 Ng et al. Table 1 Patients (a) Versus Controls (b): Demographics and Pre-VPS measurements of EI, pdfr, psfr and ADC Patient/control Age (years) Sex EI pdfr psfr ADC 1a 79 M b a 57 M b a 75 F b a 75 F b a 66 F b a 66 M b a 72 M b a 73 M b a 69 M b a patient; b age/gender-matched control; VPS ventriculoperitoneal shunting; M male; F female; EI Evans index; pdfr peak diastolic flow rate; psfr peak systolic flow rate; ADC apparent diffusion coefficient. markers (EI and FR) did not show up opposing trends for the two groups. Statistical comparison of the demographic characteristics between the responders and nonresponders (Table 5) yielded no differences in age and gender distribution. The statistical comparison of pre and postshunt measurements, expressed as an interval percentage change, between responders and nonresponders is summarized in Table 6. The percentage of total cases posting interval improvement of the various measurements (EI, FR, and ADC) for responders and nonresponders is expressed graphically in Figure 4. This reveals 100% concordance between response and decrease in ADC. Conversely ADC showed interval elevation in all nonresponders. Table 2 Statistical Comparison of Pre-VPS Measurements Between Patients and Controls Patients (n 9) Controls (n 9) P value EI Mean(SD) 0.33 (0.03) 0.27 (0.02) Range 0.28 to to Median pdfr Mean(SD) 25.1 (10.4) 7.2 (1.7) Range 10.8 to to Median psfr Mean(sd) 14.4 (11.0) 6.3 (1.7) Range 30.0 to to Median ADC Mean(SD) (2.2) 8.75 (0.99) Range 10.3 to to Median VPS ventriculo-peritoneal shunting; EI Evans index; pdfr peak diastolic flow rate; psfr peak systolic flow rate; ADC apparent diffusion coefficient. Responders showed a statistically significant decrease in ADC compared with nonresponders even after adjusting for age and gender (P 0.032). All six responders showed interval decrease compared with all the nonresponders having an elevation. Comparing the other two pre- and postshunt measurements between groups, EI posted a drop in all six (100%) responders and two of three (67%) nonresponders. FR (i.e., both pdfr and psfr) showed a decrease in four of six (67%) responders and one of three (33%) nonresponders. These differences were not statistically significant. For the nonresponders, all three patients developed SDH on their first follow-up scan, a negative finding among the group of responders (P 0.012). The caregivers of the three nonresponders gave a similar history of initial clinical improvement followed by gradual deterioration. By 3 months postoperation, the three nonresponders had stabilized and did not show further deterioration subsequently. Table 3 Pre (a) and Post (b) Shunt Measurements for Responders Patient Age (years) Sex EI pdfr psfr ADC mtii Oth Fi 1a 79 M b a 57 M b a 75 F b a 75 F b a 66 F b a 66 M b a pre-shunt measurement; b post-shunt measurement; M male; F female; EI Evans index; pdfr peak diastolic flow rate; psfr peak systolic flow rate; ADC apparent diffusion coefficient; mtii modified total improvement index; Oth Fi other findings.

6 MRI Biomarkers in Idiopathic NPH 713 Table 4 Pre (a) and Post (b) Shunt Measurements for Nonresponders Patient Age (years) Sex Ev In pdfr psfr ADC mtii Oth Fi 7a 72 M b SDH on shunt side 8a 73 M b Bilat SDH 9a 69 M b SDH on shunt side a pre-shunt measurement; b post-shunt measurement; M male; F female; EI Evans index; pdfr peak diastolic flow rate; psfr peak systolic flow rate; ADC apparent diffusion coefficient; mtii modified total improvement index; Oth Fi other findings. DISCUSSION Comparing Preshunt Measurements Between NPH Cases and Controls Diagnostic MRI Markers for NPH Comparing preshunt measurements between NPH cases and controls yielded statistically significant readings for EI (P ), pdfr (P ), and ADC (P ). Comparison for psfr yielded near statistical significance (P 0.075). Extrapolating from Table 2, top normal readings for EI, pdfr, psfr, and ADC would be 0.3, 10 ml/min, 9.0 ml/min, and x 10 4 mm 2 /s. For patients presenting with signs and symptoms of NPH, MRI readings greater than these would add further weight to the diagnosis. Disparity in Flow Rate Readings In a prior study by this group (21), the upper limit of normal pdfr and psfr was established to be 13.3 ml/min and 10.6 ml/min, respectively. This compared favorably with those reported by other investigators (10). In the latter study, all 18 patients with NPH posted elevated flow rates, with pdfr consistently higher than psfr. In our study, this pattern was observed, with only the pdfr being statistically significant compared with ageand gender-matched controls. We postulate two reasons for the persistent elevation of pdfr over psfr, as well as the statistical significance of the pdfr. The cerebral aqueduct connects a larger chamber (the third ventricle) to a smaller (fourth ventricle). This volume disparity appears to evoke higher flow during diastole (when there is retrograde flow from a smaller chamber to a larger one) than systole. This difference seems more pronounced in NPH cases than in normals. There has also been a recent suggestion that in NPH, reduced Table 5 Demographic Characteristics of Responders and Nonresponders Responders (n 6) Nonresponders (n 3) P value Total (n 9) Age Mean (SD) 69.7 (8.1) 71.3 (2.1) 70.2 (6.6) Range to to 79 Median Gender Male 3 (50%) 3 (100%) (66.7%) Female 3 (50%) 0 (0%) 3 (33.3%) compliance alters CSF and venous outflow and arterial inflow such that a smaller increase in systolic volume is posted (23,24). Comparing Pre- and Postshunt Measurements Between NPH Cases Value of ADC: Major Pathway for Reversal The interval decrease of ADC in all six (100%) responders, with concomitant interval elevation in all three (100%) nonresponders showed strong statistical significance (P 0.032). This finding of an interval ADC decrease as one of two statistically significant discriminants between shunt responders and nonresponders lends support to our hypothesis of water accumulation as a major determinant of the patient s symptoms. There are many causes for elevated ADC, chief of which are related to gliosis. All changes other than water retention are permanent and would not be expected to reverse on shunting. The converse was also true, that in instances of absent clinical recovery, the ADC showed an interval increase. This concurs with a recent report of elevated ADC in NPH patients (12), where Bradley et al presented an elegant explanation for excess water accumulating within the extracellular space in NPH patients as the cause for the elevated ADC. This is also supported by others alluding to an alternative, transventricular route of CSF absorption in NPH (11,25,26). We postulate that the main pathway that shunting reverses the symptoms of NPH is by facilitating the clearance of CSF, not so much from the ventricles but from the extracellular space of the brain. SDH In addition to interval elevation of ADC, the other statistically significant discriminator between shunt responders and nonresponders was the presence of SDH. Reduced stiffness modulus of brain parenchyma as a main cause of NPH has been postulated by some authors (27,28). Others have supported this theory by proposing that the main purpose of ventricular shunting is to increase the modulus by providing additional capacitance to the ventricles (29). We propose that the initial response to shunting in the nonresponders was due to the additional capacitance provided by the shunt. The reversal of improvement, it then follows, may be related to the subdural collection. The subdural collection could have resulted in an increase in intra-

7 714 Ng et al. Table 6 Comparison of Measurements Expressed as Percentage Change Pre- and Post-shunt, Between Responders and Nonresponders Responders (n 6) Nonresponders (n 3) Unadjusted P value Adjusted P value a % change EI Mean(SD) 5.4 (2.2) 7.8 (19.6) Range 9.1 to to 14.3 Median % change pdfr Mean(SD) 6.76 (77.4) (26.1) Range 85.0 to to Median % change psfr Mean(SD) (116.6) (36.1) Range to to 13.3 Median % change ADC Mean(SD) 7.2 (5.0) 7.7 (4.0) Range 12.3 to to 10.0 Median a Adjusting for age and gender. EI Evans index; pdfr peak diastolic flow rate; psfr peak systolic flow rate; ADC apparent diffusion coefficient. cranial pressure (ICP), compounding the drop in capacitance. It would appear that, as the collection regressed, the ICP normalized while the capacitance drop persisted. The results here would require further validation with a larger number of similar cases. Comparing EI and FRs between responders and nonresponders, the difference was not statistically significant. We believe, based on their postoperative history, that the positive predictive value of these markers was compromised by the unfortunate subdural collections. Figure 4. Percentage of total cases posting interval improvement of the various measurements (Evans index, [EI], aqueductal flow rate [FR], and apparent diffusion coefficient [ADC]) for responders and nonresponders. Note the 100% concordance between response and decrease in ADC. Conversely ADC showed interval elevation in all nonresponders.

8 MRI Biomarkers in Idiopathic NPH 715 The limitations of this study should be noted. The sample size, particularly of the shunt nonresponders, was small. It should be noted that both landmark studies comparing pre- and postshunt MRI measurements were characterized by small sample sizes of between 6 and 11, respectively (10,30). Also, there were no falsepositive cases in our study. This could have been skewed by our surgeons dependence on the pfr as a key determinant for VPS. There were also no false-positive cases without SDH, a finding that lends further credence to both pfr as a marker of shunt response and of SDH as a determinant for shunt nonresponse. While ADC posted a decrease in all six shunt responders, the decrease was small in two cases. The value of EI, pdfr, psfr, and ADC as determinants for shunt response was not established. SDH is a major determinant for shunt nonresponse and is a primary confound for assessing the utility of MRI findings to predict favorable outcomes. The inconsistent postoperative changes in pfr and EI as well as the consistent postoperative decline in ADC support our belief that water accumulation in the cerebrum, rather than ventriculomegaly or elevated FRs, is the major cause for the symptoms of NPH. We believe that shunting facilitates clearance of this excess. A postshunt decrease in ADC augurs well for the patient. Conversely, a subsequent elevation of ADC may be an early marker of shunt malfunction. The strong correlate between SDH and shunt nonresponse raises suspicion of decreased intracerebral compliance as the other major cause for the symptoms of NPH. By facilitating the clearance of this excess water, shunting adds capacitance to the cerebral hemispheres and ventricles. This in turn improves the overall compliance of the cerebrum and, hence, reverses the symptoms of NPH. SDH is a sign of overdrainage and results in raised intracranial pressure, which reverses this improvement of compliance and prevents recovery of symptoms. REFERENCES 1. Adams RD, Fisher CM, Hakim S, Ojemann RG, Sweet WH. Symptomatic occult hydrocephalus with normal cerebrospinal fluid pressure: a treatable syndrome. N Engl J Med 1965;273: Vanneste JAL. Diagnosis and management of normal-pressure hydrocephalus. J Neurol 2000;247: Evans WA. An encephalographic ratio for estimating ventricular enlargement and cerebral atrophy. Arch Neurol Psychiatry 1942; 47: Black PM. Idiopathic normal pressure hydrocephalus. Results of shunting In 62 patients. J Neurosurg 1980;52: Symon L, Dorsch NWC. Use of long-term intracranial pressure measurement to assess hydrocephalic patients prior to shunt surgery. J Neurosurg 1975;42: Pettorossi VE, Di Rocco C, Mancinelli R, Caldareli M, Velardi F. Communicating hydrocephalus induced by mechanically increased amplitude of the intraventricular cerebrospinal fluid pulse pressure: rationale and method. Exp Neurol 1978;59: Borgesen SE, Gjerris F. The predictive value of conductance to outflow of cerebrospinal fluid in normal pressure hydrocephalus. Brain 1992;105: Raftopoulos C, Deleval J, Chaskis C, et al. Cognitive recovery in idiopathic normal pressure hydrocephalus: a prospective study. Neurosurgery 1994;35: Bradley WG. Normal pressure hydrocephalus: new concepts on etiology and diagnosis. AJNR Am J Neuroradiol 2000;21: Parkkola RK, Komu ME, Kotiainen EM, Valtonen SO, Thomsen C, Gideon P. Cerebrospinal fluid flow in patients with dilated ventricles studied with MR imaging. Eur Radiol 2000;10: Poca MA, Sahuguillo J, Busto M, et al. Agreement between CSF flow dynamics in MRI and ICP monitoring in the diagnosis of normal pressure hydrocephalus. Sensitivity and specificity of CSF dynamics to predict outcome. Acta Neurochir Suppl 2002;81: Bradley WG, Bahl G, Alksne JF. Idiopathic normal pressure hydrocephalus may be a two hit disease: benign external hydrocephalus in infancy followed by deep white matter ischemia in late adulthood. J Magn Reson Imaging 2006;24: Enzmann DR, Pelc NJ. Normal flow patterns of intracranial and spinal cerebrospinal fluid defined with phase contrast cine MR imaging. Radiology 1991;178: Nitz WR, Bradley WG, Watanabe AS, et al. Flow dynamics of cerebrospinal fluid: assessment with PC velocity MRI performed with retrospective cardiac gating. Radiology 1992;183: Gideon P, Stahlberg E, Thomsen C, et al. Cerebrospinal fluid flow and production in patients with normal pressure hydrocephalus studied by MRI. Neuroradiology 1994;36: Feinberg DA, Mark AS. Human brain motion and cerebrospinal fluid circulation demonstrated with MR velocity imaging. Radiology 1987;163: Luetmer PH, Huston J, Friedman JA, et al. Measurement of cerebrospinal fluid flow at the cerebral aqueduct by use of phasecontrast magnetic resonance imaging: technique validation and utility in diagnosing idiopathic normal pressure hydrocephalus. Neurosurgery 2002;50: Bradley WG, Scalzo D, Queralt J et al. Normal pressure hydrocephalus: evaluation with cerebrospinal fluid flow measurements at MR imaging. Radiology 1996;198: Ng SES, Kwang AL, Khoo JCM, Hoe JWM, Kwok RK. Cerebrospinal fluid flow rate: the role of phase contrast MRI in normal pressure hydrocephalus. Neuroradiology 1999;41(S1): Chun T, Filippi CG, Zimmerman RD, Ulug AM. Diffusion changes in the aging human brain. AJNR Am J Neuroradiol 2000;21: Schwarcz A, Ursprung Z, Berente Z, et al. In vivo brain edema classification: new insight offered by large b-value diffusionweighted MR imaging. J Magn Reson Imaging 2007;25: Krauss JK, Droste DW, Vach W, et al. Cerebrospinal fluid shunting in idiopathic normal pressure hydrocephalus of the elderly: effect of periventricular and deep white matter lesions. Neurosurgery 1996; 39: Miyati T, Mase M, Kasai H, et al. Noninvasive MRI assessment of intracranial compliance in idiopathic normal pressure hydrocephalus. J Magn Reson Imaging 2007;26: Alperin N, Sivaramakrishnan A, Lichtor T. Magnetic resonance imaging-based measurements of cerebrospinal fluid and blood flow as indicators of intracranial compliance in patients with Chiari malformation. J Neurosurg 2005;103: Bering EA, Sato O. Hydrocephalus: changes in formation and absorption of cerebrospinal fluid within the cerebral ventricles. J Neurosurg 1963;20: Strecker EP, Scheffel U, Kelley JE, James AE Jr. Cerebrospinal fluid absorption in communicating hydrocephalus. Evaluation of transfer of radioactive albumin from subarachnoid space to plasma. Neurology 1973;23: Peňa A, Harris NG, Bolton MD, Czosnyka M, Pickard JD. Communicating hydrocephalus: the biomechanics of progressive ventricular enlargement revisited. Acta Neurochir Suppl 2002;81: Pang D, Altschuler E. Low-pressure hydrocephalus state and viscoelastic alterations in the brain. Neurosurgery 1994;35: Skylar FH, Lindler ML. The role of the pressure volume relationship of brain elasticity in the in the mechanics and treatment of hydrocephalus. In: Shapiro K, Marmarou A, Portnoy H, editors. Hydrocephalus. New York: Raven Press; p Corkill RG, Garnett MR, Blamire AM, et al. Multi-modal MRI in Normal Pressure Hydrocephalus identifies preoperative haemodynamic and diffusion coefficient changes in normal appearing white matter correlating with surgical outcome. Clin Neurol Neurosurg 2003;105: