Article
Computer Science, Artificial Intelligence
Shunjie Dong, Zixuan Pan, Yu Fu, Qianqian Yang, Yuanxue Gao, Tianbai Yu, Yiyu Shi, Cheng Zhuo
Summary: In this paper, an enhanced Deformable U-Net (DeU-Net) is proposed for cardiac MRI segmentation. The DeU-Net consists of three modules: Temporal Deformable Aggregation Module (TDAM), Enhanced Deformable Attention Network (EDAN), and Probabilistic Noise Correction Module (PNCM). Experimental results demonstrate the state-of-the-art performance of DeU-Net on the Extended ACDC dataset.
MEDICAL IMAGE ANALYSIS
(2022)
Article
Biology
Elham Avard, Isaac Shiri, Ghasem Hajianfar, Hamid Abdollahi, Kiara Rezaei Kalantari, Golnaz Houshmand, Kianosh Kasani, Ahmad Bitarafan-rajabi, Mohammad Reza Deevband, Mehrdad Oveisi, Habib Zaidi
Summary: This study demonstrates that radiomics analysis on non-contrast Cine-CMR images can accurately differentiate between myocardial infarction (MI) and normal tissue, potentially serving as an alternative diagnostic method.
COMPUTERS IN BIOLOGY AND MEDICINE
(2022)
Article
Engineering, Biomedical
Abderazzak Ammar, Omar Bouattane, Mohamed Youssfi
Summary: This study proposes an automated pipeline for cardiac MRI segmentation and diagnosis using fully convolutional neural networks, achieving nearly state-of-the-art accuracy for both segmentation and disease classification challenges.
COMPUTERIZED MEDICAL IMAGING AND GRAPHICS
(2021)
Article
Engineering, Biomedical
Abderazzak Ammar, Omar Bouattane, Mohamed Youssfi
Summary: Cardiac cine MRI, using fully convolutional neural networks, can accurately segment heart structures and predict diseases. An automated pipeline for heart segmentation and diagnosis was proposed, achieving nearly state-of-the-art accuracy in the segmentation contest and disease classification challenge of the ACDC challenge.
COMPUTERIZED MEDICAL IMAGING AND GRAPHICS
(2021)
Article
Engineering, Biomedical
Abderazzak Ammar, Omar Bouattane, Mohamed Youssfi
Summary: This study proposes an automated pipeline for cardiac segmentation and diagnosis using MRI images and deep learning techniques. By combining three classifiers, the system achieved a high accuracy for heart disease classification on unseen data.
COMPUTERIZED MEDICAL IMAGING AND GRAPHICS
(2021)
Article
Engineering, Biomedical
Abderazzak Ammar, Omar Bouattane, Mohamed Youssfi
Summary: This study utilizes cardiac cine magnetic resonance imaging for cardiac structure segmentation and disease prediction, achieving impressive results in medical imaging competitions. The automated pipeline proposed in the research uses deep learning for cardiac structure segmentation and disease diagnosis.
COMPUTERIZED MEDICAL IMAGING AND GRAPHICS
(2021)
Article
Engineering, Biomedical
D. Chen, T. Schaeffter, C. Kolbitsch, A. Kofler
Summary: By leveraging a transfer-learning approach using 2D spatio-temporal profiles as training data, a CNN method can be successfully applied to cardiac cine MRI image reconstruction, showing comparable or even superior performance compared to CNNs trained on actual cine MR images. This method effectively overcomes challenges related to data quality and quantity limitations in fast dynamic processes.
PHYSICS IN MEDICINE AND BIOLOGY
(2021)
Article
Computer Science, Information Systems
Hela Elmannai, Hager Saleh, Abeer D. Algarni, Ibrahim Mashal, Kyung Sup Kwak, Shaker El-Sappagh, Sherif Mostafa
Summary: This study proposes a stacking ensemble model based on Convolutional Neural Network (CNN) for identifying the risk of myocardial infarction. The proposed model achieves remarkable results in accuracy and performance compared to other methods, based on the evaluation of two ECG heartbeat signals datasets.
Article
Computer Science, Artificial Intelligence
Yan Xia, Nishant Ravikumar, John P. Greenwood, Stefan Neubauer, Steffen E. Petersen, Alejandro F. Frangi
Summary: The study presents a novel super-resolution algorithm based on conditional generative adversarial networks for generating high-quality cardiac MR images, which benefits subsequent image analyses and demonstrates superior performance in experiments.
MEDICAL IMAGE ANALYSIS
(2021)
Article
Computer Science, Interdisciplinary Applications
Qing Lyu, Hongming Shan, Yibin Xie, Alan C. Kwan, Yuka Otaki, Keiichiro Kuronuma, Debiao Li, Ge Wang
Summary: In this study, a novel deep learning approach for reducing motion artifacts in cardiac MRI was proposed using recurrent generative adversarial networks. The model demonstrated improved performance in image quality and temporal resolution compared to existing state-of-the-art methods through extensive experiments.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2021)
Article
Radiology, Nuclear Medicine & Medical Imaging
Mohamad Abdi, Kenneth C. Bilchick, Frederick H. Epstein
Summary: This study introduces a model to describe the effects of respiratory motion in DENSE and develops a deep convolutional neural network for respiratory motion compensation. The model is validated using phantom experiments and simulations, and used to train the DENSE-RESP-NET for motion correction. The corrected DENSE provides reliable strain measurements for myocardial parameters.
MAGNETIC RESONANCE IN MEDICINE
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Kaiyue Diao, Hong-qing Liang, Hong-kun Yin, Ming-jing Yuan, Min Gu, Peng-xin Yu, Sen He, Jiayu Sun, Bin Song, Kang Li, Yong He
Summary: A fully automatic framework for the diagnosis of left ventricular hypertrophy (LVH) was developed using cardiac cine images. Model 3, based on the ventricle mask, achieved the best performance with an overall accuracy of 77.4% in the external test dataset, as well as high AUCs for binary classification tasks.
INSIGHTS INTO IMAGING
(2023)
Article
Chemistry, Analytical
Soichiro Inomata, Takaaki Yoshimura, Minghui Tang, Shota Ichikawa, Hiroyuki Sugimori
Summary: This study aims to estimate cardiac function indices by learning short-axis images and known left and right ventricular ejection fractions, confirming accuracy and capturing each index as a feature. A regression model using 3D-CNN was built and trained with a dataset of publicly available short-axis cine images, with accuracy assessed using five-fold cross-validation. The mean correlation coefficient, MAE, and RMSE for left ventricular ejection fraction were 0.80, 9.41, and 12.26 respectively, while for right ventricular ejection fraction they were 0.56, 11.35, and 14.95. Left ventricular ejection fraction was estimated more accurately, capturing left ventricular systolic function as a feature.
Article
Radiology, Nuclear Medicine & Medical Imaging
Bing-Hua Chen, Chong-Wen Wu, Dong-Aolei An, Ji-Lei Zhang, Yi-Hong Zhang, Ling-Zhan Yu, Kennedy Watson, Luke Wesemann, Jiani Hu, Wei-Bo Chen, Jian-Rong Xu, Lei Zhao, ChaoLu Feng, Meng Jiang, Jun Pu, Lian-Ming Wu
Summary: The study aimed to develop a DCNN model that integrates multidimensional CMR data to accurately identify LV paradoxical pulsation after reperfusion by primary percutaneous coronary intervention. The DCNN model showed high accuracy in identifying paradoxical pulsation, and the 2.5D multiview model performed better than the 3D model. The discrimination performance of the DCNN model was superior to that of trained physicians.
EUROPEAN RADIOLOGY
(2023)
Article
Physics, Multidisciplinary
Shuihua Wang, Ahmed M. S. E. K. Abdelaty, Kelly Parke, Jayanth Ranjit Arnold, Gerry P. McCann, Ivan Y. Tyukin
Summary: Myocardial infarction (MI) is a condition where the heart does not receive enough blood due to the blockage of a coronary artery. This study introduces an end-to-end automated system (MyI-Net) for the detection and quantification of MI in magnetic resonance images. The system utilizes four processing stages to extract features and perform image segmentation, showing improved performance compared to other methods.
Article
Computer Science, Artificial Intelligence
Xi Zhou, Qinghao Ye, Xiaolin Yang, Jiakun Chen, Haiqin Ma, Jun Xia, Javier Del Ser, Guang Yang
Summary: Based on CT and MRI images, machine learning methods were used to establish a multimodal ventricle segmentation method for efficient and accurate measurement of ventricular volume in NPH patients. The results showed high reliability in both CT and MRI images, indicating the potential of this method for assisting clinicians in understanding the ventricle situation of NPH patients.
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Engineering, Multidisciplinary
Xiaofei Xue, Xiujian Liu, Zhifan Gao, Rui Wang, Lei Xu, Dhanjoo Ghista, Heye Zhang
Summary: This study proposes a personalized coronary flow model, CTPV, based on dynamic stress computed tomographic perfusion, to achieve a noninvasive estimation of coronary outlet resistance and FFRCT. The CTPV model directly quantifies the total hyperemic coronary blood flow using myocardial blood flow obtained from CTP. It then distributes the outlet coronary blood flow based on myocardial perfusion territories corresponding to coronary artery branches. The experimental results demonstrate that the CTPV model accurately estimates FFRCT compared to the CTPD and LVMD models.
COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Hao Li, Yang Nan, Javier Del Ser, Guang Yang
Summary: In this paper, we propose a region-based Evidential Deep Learning (EDL) segmentation framework that can generate reliable uncertainty maps and accurate segmentation results, which is robust to noise and image corruption.
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Boyan Wang, Xuegang Hu, Jinglin Zhang, Chenchu Xu, Zhifan Gao
Summary: Mammography screening is an important application of intelligent IoT, which can reduce breast cancer mortality rate through AI-driven remote assistant diagnosis. This article proposes a novel framework called MLT-UniCaps, which utilizes modified capsule neural network and multicenter mammograms to improve the effectiveness and robustness of mammography screening. Experimental results show that MLT-UniCaps achieves 90.1% overall classification accuracy in single-center trials and 73.8% overall F1 score in multicenter trials.
IEEE INTERNET OF THINGS JOURNAL
(2023)
Article
Computer Science, Interdisciplinary Applications
Hongwei Zhang, Zhifan Gao, Dong Zhang, William Kongto Hau, Heye Zhang
Summary: Main coronary segmentation is crucial for the computer-aided diagnosis and treatment of coronary disease. However, it faces challenges at different image granularities, including semantic confusion, low contrast, and local ambiguity. Traditional feature-based methods lack semantic relationship and cannot distinguish vessels, while existing deep learning methods have issues in capturing long-distance relationships, handling interference, and preserving boundary details. To address these challenges, the progressive perception learning (PPL) framework is proposed, which includes context, interference, and boundary perception modules. Extensive experiments demonstrate the effectiveness of PPL, outperforming thirteen state-of-the-art methods.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2023)
Article
Computer Science, Artificial Intelligence
Chengjin Yu, Shuang Li, Dhanjoo Ghista, Zhifan Gao, Heye Zhang, Javier Del Ser, Lin Xu
Summary: Most existing methods for cardiac echocardiography segmentation require a large number of labeled data, which is time-consuming and laborious for physicians. In this study, we propose a fusion method of multi-level and multi-type self-generated knowledge to address this challenge. We extract multi-level sub-anatomical structure information from ultrasound images using a superpixel method, and then fuse various types of information generated by multiple pretext tasks. The experimental results demonstrate the effectiveness of our method in echocardiography segmentation task, achieving comparable performance to fully supervised methods without requiring a high amount of labeled data.
INFORMATION FUSION
(2023)
Article
Computer Science, Theory & Methods
Dong Zhang, Heye Zhang, Hongwei Zhang, Lei Xu, Jinglin Zhang, Zhifan Gao
Summary: The structural and functional analysis of coronary arteries in XRA images is crucial for intraoperative treatment of coronary artery disease. AngioSFA, an efficient framework, integrates structural analysis with functional analysis by utilizing DTSN for structural analysis and DSLN for functional analysis. Experimental results show that AngioSFA achieves high accuracy in both structural and functional analysis, indicating its great potential in the intraoperative treatment of coronary artery disease.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2023)
Article
Computer Science, Artificial Intelligence
Saidi Guo, Xiujian Liu, Heye Zhang, Qixin Lin, Lei Xu, Changzheng Shi, Zhifan Gao, Antonella Guzzo, Giancarlo Fortino
Summary: In this paper, we propose the causal knowledge fusion (CKF) framework to solve the challenge of 3D cross-modality cardiac image segmentation. The CKF explores causal intervention to obtain the anatomical factor and discards the modality factor, improving the information fusion and spatial learning ability. Experimental results show that the CKF is effective and superior to state-of-the-art segmentation methods.
INFORMATION FUSION
(2023)
Article
Computer Science, Information Systems
Zeyu Tang, Yang Nan, Simon Walsh, Guang Yang
Summary: This paper proposes an adversarial-based refinement network that takes preliminary segmentation and original CT images as input and outputs a refined mask of the airway structure. Experimental results demonstrate the promising performance of the method in terms of accuracy and robustness in airway segmentation, effectively detecting discontinuities and missing bronchioles.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2023)
Article
Computer Science, Information Systems
Xiaodan Xing, Javier Del Ser, Yinzhe Wu, Yang Li, Jun Xia, Lei Xu, David Firmin, Peter Gatehouse, Guang Yang
Summary: Synthetic digital twins based on medical data accelerate the acquisition, labelling, and decision making procedure in digital healthcare. This study proposes a hybrid deep learning network for synthetic cardiac data, which can successfully synthesize high temporal resolution myocardial velocity mapping data.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2023)
Article
Automation & Control Systems
Yang Li, Yue Zhang, Jing-Yu Liu, Kang Wang, Kai Zhang, Gen-Sheng Zhang, Xiao-Feng Liao, Guang Yang
Summary: In this article, a global transformer (GT) and dual local attention (DLA) network via deep-shallow hierarchical feature fusion (GT-DLA-dsHFF) are investigated to improve the segmentation performance of retinal blood vessels. Experimental results demonstrate the superior performance of the proposed method in segmenting fine vessels, and the efficacy of the three proposed modules is verified.
IEEE TRANSACTIONS ON CYBERNETICS
(2023)
Article
Engineering, Electrical & Electronic
Xin Zhao, Tongming Wang, Jingsong Chen, Bingrun Jiang, Haotian Li, Nan Zhang, Guang Yang, Senchun Chai
Summary: This study proposes a novel contrastive learning strategy that leverages the relative position differences between image slices, combined with global and local features, to address the challenge of generating positive and negative data pairs for medical image segmentation. By employing a two-dimensional fully connected conditional random field for iterative optimization, segmentation accuracy is enhanced and isolated mis-segmented regions are reduced. Experimental results demonstrate that this method outperforms existing semi-supervised and self-supervised techniques in medical segmentation tasks with limited annotated samples.
INTERNATIONAL JOURNAL OF IMAGING SYSTEMS AND TECHNOLOGY
(2023)
Article
Medicine, General & Internal
Yongwon Cho, Soojung Park, Sung Ho Hwang, Minseok Ko, Do-Sun Lim, Cheol Woong Yu, Seong-Mi Park, Mi-Na Kim, Yu-Whan Oh, Guang Yang
Summary: This study proposes a deep learning architecture that can automatically detect the complex structure of the aortic annulus plane for transcatheter aortic valve replacement (TAVR) using cardiac computed tomography (CT). The performance of the proposed model, ADPANet, was evaluated and found to outperform other convolutional neural networks in terms of detecting the aortic annulus plane.
JOURNAL OF KOREAN MEDICAL SCIENCE
(2023)
Correction
Radiology, Nuclear Medicine & Medical Imaging
Zhen Zhou, Yifeng Gao, Weiwei Zhang, Kairui Bo, Nan Zhang, Hui Wang, Rui Wang, Zhiqiang Du, David Firmin, Guang Yang, Heye Zhang, Lei Xu
EUROPEAN RADIOLOGY
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Zhen Zhou, Yifeng Gao, Weiwei Zhang, Kairui Bo, Nan Zhang, Hui Wang, Rui Wang, Zhiqiang Du, David Firmin, Guang Yang, Heye Zhang, Lei Xu
Summary: This study successfully reduces the contrast medium dose of full aortic CT angiography while maintaining image quality and diagnostic accuracy using the Au-CycleGAN algorithm.
EUROPEAN RADIOLOGY
(2023)