Premium
Discrimination between patients with mild A lzheimer's disease and healthy subjects based on cerebral blood flow images of the lateral views in xenon‐enhanced computed tomography
Author(s) -
Sase Shigeru,
Yamamoto Homaro,
Kawashima Ena,
Tan Xin,
Sawa Yutaka
Publication year - 2018
Publication title -
psychogeriatrics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.647
H-Index - 32
eISSN - 1479-8301
pISSN - 1346-3500
DOI - 10.1111/psyg.12281
Subject(s) - cerebral blood flow , volunteer , region of interest , medicine , nuclear medicine , receiver operating characteristic , radiology , cardiology , biology , agronomy
Background Quantitative cerebral blood flow ( CBF ) measurement is expected to help early detection of functional abnormalities caused by A lzheimer's disease ( AD ) and enable AD treatment to begin in its early stages. Recently, a technique of layer analysis was reported that allowed CBF to be analyzed from the outer to inner layers of the brain. The aim of this work was to develop methods for discriminating between patients with mild AD and healthy subjects based on CBF images of the lateral views created with the layer analysis technique in xenon‐enhanced computed tomography. Methods Xenon‐enhanced computed tomography using a wide‐volume CT was performed on 17 patients with mild AD aged 75 or older and on 15 healthy age‐matched volunteers. For each subject, we created CBF images of the right and left lateral views with a depth of 10–15 mm from the surface of the brain. Ten circular regions of interest ( ROI ) were placed on each image, and CBF was calculated for each ROI . We determined discriminant ROI that had CBF that could be used to differentiate between the AD and volunteer groups. AD patients’ CBF range (mean − SD to mean + SD ) and healthy volunteers’ CBF range (mean − SD to mean + SD ) were obtained for each ROI . Receiver–operator curves were created to identify patients with AD for each of the discriminant ROI and for the AD patients’ and healthy volunteers’ CBF ranges. Results We selected an ROI on both the right and left temporal lobes as the discriminant ROI . Areas under the receiver–operator curve were 93.3% using the ROI on the right temporal lobe, 95.3% using the ROI on the left temporal lobe, and 92.4% using the AD patients’ and healthy volunteers’ CBF ranges. Conclusions We could effectively discriminate between patients with mild AD and healthy subjects using ROI placed on CBF images of the lateral views in xenon‐enhanced computed tomography.