Patients
The subjects were iNPH patients who visited Yamagata University Hospital and were diagnosed based on the guideline diagnostic criteria [16] between March 2017 and April 2021. Of these, 25 patients with iNPH who were assessed using 123I-FP-CIT SPECT before shunt surgery or the tap test were included in this study. Furthermore, to establish the diagnosis of iNPH more reliably, only patients with an Evans index greater than 0.3 were included. The iNPH stage composition consisted of four possible cases, six probable cases, and 15 definite cases. All subjects in this study underwent 99mTc-ethyl cysteinate dimer SPECT and 123I-metaiodobenzylguanidine (MIBG) myocardial scintigraphy, eliminating as much as possible DLB and PD, which are characterized by decreased blood flow in the occipital lobe [17] and/or decreased cardiac sympathetic function [17, 18]. Demographic and clinical data of the subjects are shown in Table 1.
Single-photon emission computed tomography imaging
We acquired 123I-FP-CIT SPECT/computed tomography (CT) images over 28 min using a Symbia T2 with a rotating, dual-detector gamma camera (Siemens Healthineers, Erlangen, Germany) and a low- to medium-energy general-purpose collimator (Siemens Healthineers), with 360° continuous rotation (7.0 min/rotation × 4 rotations). Supine patients at rest with their eyes closed were intravenously injected with 123I-FP-CIT (167 MBq), and SPECT/CT images were acquired for 3 h under the following conditions: magnification, 1.45; matrix, 128 × 128 (3.3 mm/pixel); main window, 159 ± 12.0 keV; and sub-window, 8%.
A Gaussian post-processing filter with a full width half maximum of 6.6 mm, six subsets, and eight iterations was applied for ordered subset expectation maximization reconstruction with CT-based attenuation correction, a multi-energy window, and collimator aperture correction for scatter correction.
Specific binding ratio (SBR) calculation
Volumes of interest were placed on the striatum and BG regions determined using the Tossici-Bolt method [6] (Figure 1). We obtained a highly reproducible quantitative index using a 44-mm-thick tomographic image centered on the striatum along with a large region of interest.
SBR was calculated as follows:
CSF area mask correction
The CSF area mask correction creates a standard normal distribution for each VOI in the striatal and BG regions and excludes the low accumulation of the set standard deviation (SD) from the SBR calculation. The SD threshold was set to an average of −1.0 SD [7].
Changes in SBR before and after CSF area mask correction
The SBR can be calculated using DaTView analysis (Nihon Medi-Physics, Tokyo, Japan), with differences between facilities corrected by applying the formula derived from the striatal phantom using SPECT/CT, y = 0.9369x + 0.0404 (19). The SBRs with and without CSF area mask correction were calculated, respectively, and changes in the quantitative values were verified.
Verification of striatum and background volume of interest (BG VOI) volumes before and after CSF area mask correction
Removing a larger volume via CSF area mask correction results in a higher VOI count. Therefore, we hypothesized that the change in SBR was dependent on whether the volume removed by CSF area mask correction was larger in the striatal VOI or BG VOI. In other words, it was estimated that the SBR decreased when the volume removed by the CSF area mask correction in the striatum VOI was larger than the volume removed by the CSF area mask correction in the BG VOI. Moreover, it was estimated that the SBR increased when the volume removed by the CSF area mask correction in the striatum VOI was smaller than the volume removed by the CSF area mask correction in the BG VOI. The number of voxels in each VOI before and after CSF area mask correction was extracted from DaTView (Nihon Medi-Physics, Tokyo, Japan). The number of voxels after was subtracted from that before CSF area mask correction, and the volume removed by the CSF area mask correction was calculated. The volumes removed from the VOIs were compared to verify their effect on the SBR.
Comparison of SBR before and after CSF area mask correction with a database of age-matched healthy controls
A database of Japanese healthy controls shows the upper and lower limits of the 95% prediction interval for each age [19]. We compared the SBR before and after CSF area mask correction with the normal SBR obtained from these controls.
Ethical statements
This study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethical Review Committee of Yamagata University Faculty of Medicine for epidemiological research (approval number: 2021-89), and the requirement for obtaining informed consent from the patients was waived because this study was retrospective design by the Ethical Review Committee of Yamagata University Faculty of Medicine (approval number: 2021-89).