A deep learning approach for quantifying CT perfusion parameters in stroke

doi:10.1088/2057-1976/adc9b6

	A deep learning approach for quantifying CT perfusion parameters in stroke
	Zeng, Wanning1 ; Li, Yang 2; Zhang, Jeff L.1 ; Chen, Tong 2; Wu, Ke 2; Zong, Xiaopeng1,3,4
	2025-05-30
发表期刊	BIOMEDICAL PHYSICS AND ENGINEERING EXPRESS (IF:1.3[JCR-2023],1.3[5-Year])
ISSN	2057-1976
EISSN	2057-1976
卷号	11 期号:3
发表状态	已发表
DOI	10.1088/2057-1976/adc9b6
摘要	Objective. Computed tomography perfusion (CTP) imaging is widely used for assessing acute ischemic stroke. However, conventional methods for quantifying CTP images, such as singular value decomposition (SVD), often lead to oscillations in the estimated residue function and underestimation of tissue perfusion. In addition, the use of global arterial input function (AIF) potentially leads to erroneous parameter estimates. We aim to develop a method for accurately estimating physiological parameters from CTP images. Approach. We introduced a Transformer-based network to learn voxel-wise temporal features of CTP images. With global AIF and concentration time curve (CTC) of brain tissue as inputs, the network estimated local AIF and flow-scaled residue function. The derived parameters, including cerebral blood flow (CBF) and bolus arrival delay (BAD), were validated on both simulated and patient data (ISLES18 dataset), and were compared with multiple SVD-based methods, including standard SVD (sSVD), block-circulant SVD (cSVD) and oscillation-index SVD (oSVD). Main results. On data simulating multiple scenarios, local AIF estimated by the proposed method correlated with true AIF with a coefficient of 0.97 ± 0.04 (P −1, and estimated BAD with a mean error of 0.51 s; the latter two errors were significantly lower than those of the SVD-based methods (P −1 or 39.33% and 8.55 ml/100 g min−1 or 57.73% (P © 2025 The Author(s). Published by IOP Publishing Ltd.
关键词	Brain Physiological models Acute ischemic stroke Arterial input function Cerebral blood flow CT perfusion imaging Deep learning Patient data Perfusion images Residue functions Singular values Value decomposition
收录类别	EI
语种	英语
出版者	Institute of Physics
EI入藏号	20251718275077
EI主题词	Mean square error
EI分类号	101.1 Biomedical Engineering ; 1202.2 Mathematical Statistics
原始文献类型	Journal article (JA)
文献类型	期刊论文
条目标识符	https://kms.shanghaitech.edu.cn/handle/2MSLDSTB/523914
专题	生物医学工程学院信息科学与技术学院_硕士生生物医学工程学院_PI研究组_张雷组（生医工）生物医学工程学院_PI研究组_宗小鹏组
作者单位	1.School of Biomedical Engineering, ShanghaiTech University, Shanghai, China; 2.United Imaging Healthcare Group, Shanghai, China; 3.State Key Laboratory of Advanced Medical Materials and Devices, ShanghaiTech University, Shanghai, China; 4.Shanghai Clinical Research and Trial Center, Shanghai, China
第一作者单位	生物医学工程学院
第一作者的第一单位	生物医学工程学院
推荐引用方式 GB/T 7714	Zeng, Wanning,Li, Yang,Zhang, Jeff L.,et al. A deep learning approach for quantifying CT perfusion parameters in stroke[J]. BIOMEDICAL PHYSICS AND ENGINEERING EXPRESS,2025,11(3).
APA	Zeng, Wanning,Li, Yang,Zhang, Jeff L.,Chen, Tong,Wu, Ke,&Zong, Xiaopeng.(2025).A deep learning approach for quantifying CT perfusion parameters in stroke.BIOMEDICAL PHYSICS AND ENGINEERING EXPRESS,11(3).
MLA	Zeng, Wanning,et al."A deep learning approach for quantifying CT perfusion parameters in stroke".BIOMEDICAL PHYSICS AND ENGINEERING EXPRESS 11.3(2025).