Measurement of cerebrospinal fluid flow at the cerebral aqueduct by use of phase-contrast magnetic resonance imaging: technique validation and utility in diagnosing idiopathic normal pressure hydrocephalus

We analyzed the reliability of a protocol for measuring quantitative cerebrospinal fluid (CSF) flow at the cerebral aqueduct and established the range of CSF flows in normal elderly patients, patients with Alzheimer's and other forms of dementia, and patients with idiopathic normal pressure hydrocephalus (NPH). A constant flow phantom was used to establish the accuracy of the CSF flow measurement. The clinical variability of the measurement was estimated by calculating the standard deviations and coefficients of variation of intra- and interobserver and intertrial data sets derived from three normal volunteers. A total of 236 patients were studied, including 47 normal elderly patients, 115 patients with cognitive impairment (9 with mild cognitive impairment, 46 with Alzheimer's disease, and 60 with other cognitive impairment), 31 patients in whom NPH was suspected but ultimately excluded, and 43 patients with a final clinical diagnosis of NPH. The intraobserver, interobserver, and intertrial measurement variations of 6.4, 5.4, and 8.8%, respectively, were substantially smaller than the wide variation observed among subjects. There was no statistically significant difference in flow between normal elderly patients and patients with cognitive impairment (P = 0.91). When these populations were pooled, the average flow was 8.47 ml/min (standard deviation, 4.23; range, 0.9-18.5 ml/min). The average flow rate in patients with a final clinical diagnosis of NPH was 27.4 ml/min (standard deviation, 15.3; range, 3.13-62.2 ml/min). This was significantly higher than the flow rate in each of the other three groups (all, P < 0.001). CSF flow measurements of less than 18 ml/min with a sinusoidal flow pattern are normal. CSF flow of greater than 18 ml/min suggests idiopathic NPH.