Automated Estimation of Time Activity Curve for Pulmonary Artery from Axillary Vein on Dynamic Scintigraphy by Using Three-layered Artificial Neural Network

DOI Open Access

Bibliographic Information

Other Title
  • Dynamic Scintigraphyにおける3層人工ニューラルネットワークを利用した腋窩静脈からの肺動脈時間放射能曲線の推定法

Search this article

Abstract

Graph Plot method is one of quantitative evaluation procedures for mean cerebral blood flow (mCBF) estimation. The mCBF is measured by a pair of time activity curves (TACs) of pulmonary artery and brain regions on dynamic scintigraphy, however, the pulmonary artery, as an input function, may be overlapped with other structures due to restriction of imaging conditions or patient movements. The purpose of this study was to obtain an estimated pulmonary artery TAC from axillary vein TAC by using three-layered artificial neural network (ANN). We employed 37 cases with IRB approval and the leave-one-case-out method for the evaluation of the estimation. As a result of this initial study, the mean and standard deviation of the mean absolute errors between the estimated and actual value of pulmonary artery TAC in each case were 22.1 [counts] and 8.8 [counts], respectively. The correlation coefficient between the estimated and actual value was 0.82. We concluded that the pulmonary artery TAC could be estimated from axillary vein TAC.

Journal

Related Projects

See more

Keywords

Details 詳細情報について

Report a problem

Back to top