干旱区生态环境知识资源中心(Arid): Hierarchical multi-level dynamic hyperparameter deformable image registration with convolutional neural network

Arid

DOI	10.1088/1361-6560/ad67a6
	Hierarchical multi-level dynamic hyperparameter deformable image registration with convolutional neural network
	Zhu, Zhenyu; Li, Qianqian; Wei, Ying; Song, Rui
通讯作者	Wei, Y ; Song, R
来源期刊	PHYSICS IN MEDICINE AND BIOLOGY
ISSN	0031-9155
EISSN	1361-6560
出版年	2024
卷号	69 期号:17
英文摘要	Objective. To enable the registration network to be trained only once, achieving fast regularization hyperparameter selection during the inference phase, and to improve registration accuracy and deformation field regularity. Approach. Hyperparameter tuning is an essential process for deep learning deformable image registration (DLDIR). Most DLDIR methods usually perform a large number of independent experiments to select the appropriate regularization hyperparameters, which are time-consuming and resource-consuming. To address this issue, we propose a novel dynamic hyperparameter block, which comprises a distributed mapping network, dynamic convolution, attention feature extraction layer, and instance normalization layer. The dynamic hyperparameter block encodes the input feature vectors and regularization hyperparameters into learnable feature variables and dynamic convolution parameters which changes the feature statistics of the high-dimensional features layer feature variables, respectively. In addition, the proposed method replaced the single-level structure residual blocks in LapIRN with a hierarchical multi-level architecture for the dynamic hyperparameter block in order to improve registration performance. Main results. On the OASIS dataset, the proposed method reduced the percentage of vertical bar J(phi)vertical bar <= 0 by <= %, 9.78 % and improved Dice similarity coefficient by 1.17 %, 1.17 %, compared with LapIRN and CIR, respectively. On the DIR-Lab dataset, the proposed method reduced the percentage of vertical bar J(phi)vertical bar <= 0 by 10.00 % , 5.70 % and reduced target registration error by 10.84 %, 10.05 %, compared with LapIRN and CIR, respectively. Significance. The proposed method can fast achieve the corresponding registration deformation field for arbitrary hyperparameter value during the inference phase. Extensive experiments demonstrate that the proposed method reduces training time compared to DLDIR with fixed regularization hyperparameters while outperforming the state-of-the-art registration methods concerning registration accuracy and deformation smoothness on brain dataset OASIS and lung dataset DIR-Lab.
英文关键词	regularization hyperparameters feature statistics deformable image registration dynamic convolution hierarchical multi-level architecture
类型	Article
语种	英语
收录类别	SCI-E
WOS记录号	WOS:001290619100001
WOS关键词	LANDMARK ; FRAMEWORK ; ROBUST
WOS类目	Engineering, Biomedical ; Radiology, Nuclear Medicine & Medical Imaging
WOS研究方向	Engineering ; Radiology, Nuclear Medicine & Medical Imaging
资源类型	期刊论文
条目标识符	http://119.78.100.177/qdio/handle/2XILL650/405064
推荐引用方式 GB/T 7714	Zhu, Zhenyu,Li, Qianqian,Wei, Ying,et al. Hierarchical multi-level dynamic hyperparameter deformable image registration with convolutional neural network[J],2024,69(17).
APA	Zhu, Zhenyu,Li, Qianqian,Wei, Ying,&Song, Rui.(2024).Hierarchical multi-level dynamic hyperparameter deformable image registration with convolutional neural network.PHYSICS IN MEDICINE AND BIOLOGY,69(17).
MLA	Zhu, Zhenyu,et al."Hierarchical multi-level dynamic hyperparameter deformable image registration with convolutional neural network".PHYSICS IN MEDICINE AND BIOLOGY 69.17(2024).