Temperature measurement of intracranial cerebrospinal fluid using diffusion tensor imaging after revascularization surgery in Moyamoya disease

Abstract

ObjectiveBrain temperature monitoring using a catheter thermometer has been reported to be a useful technique to predict prognosis in neurosurgery. To investigate the possibility of measuring intracranial cerebrospinal fluid temperature for postoperative monitoring in patients with Moyamoya disease (MMD) after bypass surgery. Materials and methodsThis study evaluated fifteen patients with MMD who were indicated for bypass surgery. Diffusion tensor imaging for brain thermometry were performed on a 1.5-T MR scanner. Intracranial cerebrospinal fluid temperature with/without considering the fractional anisotropy component, body temperature, C-reactive protein levels, white blood cell count, and cerebral blood flow measured by 123I-IMP single-photon emission computed tomography were obtained before surgery and 1–3 days after surgery. Pixel values considered to be signal outliers in fractional anisotropy processing were defined as cerebrospinal fluid noise index and calculated. Wilcoxon signed-rank test and effect size were performed to compare the changes before and after revascularization. Spearman's rho correlation coefficient was used to analyze the correlations between each parameter. Statistical significance was defined as p < 0.05. ResultsAll parameter values became significantly higher compared to those measured before revascularization (p < 0.01 in all cases). The effect sizes were largest for the cerebrospinal fluid temperature with fractional anisotropy processing and for C-reactive protein levels (Rank-biserial correlation = 1.0). The cerebrospinal fluid noise index and cerebrospinal fluid temperatures with fractional anisotropy processing (r = 0.84, p < 0.0001) or without fractional anisotropy processing (r = 0.95, p < 0.0001) showed highly significant positive correlations. Although no significant correlation was observed, cerebrospinal fluid temperatures with fractional anisotropy had small or moderately positive correlations with cerebral blood flow, body temperature, C-reactive protein levels, and white blood cell count (r = 0.37, 0.42, 0.41, and 0.44, respectively; p > 0.05). ConclusionOur findings suggest the possibility of postoperative monitoring for MMD patients by measuring intracranial cerebrospinal fluid temperature with fractional anisotropy processing. Intracranial cerebrospinal fluid temperature might be considered as combined response since cerebrospinal fluid, body temperature, and inflammation are equally correlated.