A 3D Deep Learning Architecture for Denoising Low-Dose CT Scans

doi:10.1007/978-3-031-82768-6_9

	A 3D Deep Learning Architecture for Denoising Low-Dose CT Scans
	Kasparian, Armen 1; Cao, Guohua2 ; Feng, Wu-Chun 1
	2025
会议录名称	LECTURE NOTES IN COMPUTER SCIENCE (INCLUDING SUBSERIES LECTURE NOTES IN ARTIFICIAL INTELLIGENCE AND LECTURE NOTES IN BIOINFORMATICS) ; 12TH INTERNATIONAL CONFERENCE ON COMPUTATIONAL ADVANCES IN BIO AND MEDICAL SCIENCES, ICCABS 2023
ISSN	0302-9743
卷号	14548 LNBI
页码	94-106
DOI	10.1007/978-3-031-82768-6_9
摘要	Low-dose computed tomography (LDCT) scans reduce the radiation dose of computed tomography (CT) scans but come at the expense of image quality. Deep-learning (DL) image denoising techniques can enhance these LDCT images to match the quality of their regular-dose CT counterparts. To achieve better denoising performance than the current state of the art, we present a novel 3D DL architecture for LDCT image denoising called 3D-DDnet. The architecture leverages the inter-slice correlation in volumetric CT scans to obtain better denoising performance and employs distributed data parallel (DDP) strategies along with transfer learning to achieve faster training. The DDP training strategy enables a scalable multi-GPU approach on Nvidia A100 GPUs, which allows the training of previously prohibitively large volumetric samples. Our results show that 3D-DDnet achieves 10% better mean square error (MSE) on LDCT scans than its 2D predecessor (i.e., 2D-DDnet). In addition, the transfer learning in 3D-DDnet leverages existing trained 2D models to "jump start" the weights and biases of our 3D DL model and reduces training time by 50% while maintaining accuracy. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.
关键词	Contrastive Learning Deep learning Federated learning Image denoising Transfer learning Computed tomography Computed tomography scan Data parallel De-noising Deep learning Distributed data Distributed data parallel Dose computed tomographies Low dose Transfer learning
会议地点	Norman, OK, United states
会议日期	December 11, 2023 - December 13, 2023
收录类别	EI
语种	英语
出版者	Springer Science and Business Media Deutschland GmbH
EI入藏号	20251018015580
EI主题词	Computerized tomography
EISSN	1611-3349
EI分类号	1101.2 Machine Learning ; 1101.2.1 Deep Learning ; 1106.3.1 Image Processing ; 716.1 Information Theory and Signal Processing ; 746 Imaging Techniques
原始文献类型	Conference article (CA)
文献类型	会议论文
条目标识符	https://kms.shanghaitech.edu.cn/handle/2MSLDSTB/503677
专题	生物医学工程学院生物医学工程学院_PI研究组_曹国华组
通讯作者	Kasparian, Armen
作者单位	1.Department of Computer Science, Virginia Tech, Blacksburg, United States; 2.School of Biomedical Engineering, ShanghaiTech University, Shanghai, China
推荐引用方式 GB/T 7714	Kasparian, Armen,Cao, Guohua,Feng, Wu-Chun. A 3D Deep Learning Architecture for Denoising Low-Dose CT Scans[C]:Springer Science and Business Media Deutschland GmbH,2025:94-106.