Analysis of the Consistency of MRI Signs and Associated Neuroimaging Rating Scales for Patients with Idiopathic Normal Pressure Hydrocephalus

Objective: To investigate the consistency of MRI signs and associated neuroimaging rating scales for patients with idiopathic normal pressure hydrocephalus (iNPH). Methods: Imaging data were retrospectively collected from 80 patients with iNPH who underwent shunt surgery at our hospital from 2018 to 2021. The patient images were first scored using the criteria of the iNPH Radscale score and disproportionately enlarged subarachnoid hydrocephalus (DESH) score. Then, one month after the initial scoring, the patient images were scored again using the “simple typical rapid-learning standardized images of the iNPH Radscale and DESH scores.” Finally, the scores from different time periods were analyzed to assess the consistency of each parameter obtained from different observers at the same time and from the same observer at different times. Results: The Evans index (a quantitative indicator) showed high consistency in both rating scales, with kappa coefficients of 0.784 (DESH scores) and 0.806 (iNPH Radscale scores), which were statistically significant. By contrast, the consistency of the qualitative indicators was generally relatively low, with the exception of the callosal angle, which had a kappa coefficient of 0.625 in the DESH scores. The rest of the kappa coefficients in the DESH scores were statistically significant at 0.442, 0.220, and 0.333 for Sylvian fissures, tight high convexity, and focal sulcal dilatation, respectively. The kappa coefficients in the iNPH Radscale scores were also statistically significant at 0.441, 0.550, 0.201, 0.301, 0.470, and 0.504 for the callosal angle, Sylvian fissures, tight high convexity, focal sulcal dilatation, temporal horn, and periventricular hyperintensities, respectively. The consistency of the indicators was generally improved after the use of the “simple typical rapid-learning standardized images of the iNPH Radscale and DESH scores.” The kappa coefficients of the Evans index, callosal angle, Sylvian fissures, tight high convexity, and focal sulcal dilatation indicators in the DESH scores were 0.812, 0.848, 0.809, 0.732, and 0.668, respectively, whereas those of the Evans index, callosal angle, Sylvian fissures, tight high convexity, focal sulcal dilatation, temporal horn, and periventricular hyperintensities indicators in the iNPH Radscale scores were 0.822, 0.610, 0.848, 0.710, 0.781, 0.616, and 0.831, respectively. Conclusion: The consistency of the iNPH Radscale and DESH scores was not good. By contrast, the “simple typical rapid-learning standardized images of the iNPH Radscale and DESH scores,” as a new measurement method, improved the consistency of the scores, reduced the influence of subjective factors, and provided a more accurate and reliable basis for the clinical diagnosis and treatment of patients with iNPH.