Improving myocardial pathology segmentation with U-Net plus plus and EfficientSeg from multi-sequence cardiac magnetic resonance images

Article Computer Science, Information Systems

Constraint-Based Unsupervised Domain Adaptation Network for Multi-Modality Cardiac Image Segmentation

Xiuquan Du et al.

Summary: This paper proposes a constraint-based unsupervised domain adaptation network for cardiac image segmentation. The network performs image translation, pseudo-labeling, and cross-domain self-supervised learning to tackle the challenges in cardiac image segmentation. Experimental results demonstrate the important assistance of this method in the clinical evaluation of unannotated cardiac datasets.

IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS (2022)

Add to Collection

Article Engineering, Electrical & Electronic

Learning multi-level structural information for small organ segmentation

Yueyun Liu et al.

Summary: This paper proposes a multi-level structural loss and a multi-branch network, achieving significant results in small organ segmentation tasks and improving segmentation accuracy.

SIGNAL PROCESSING (2022)

Add to Collection

Article Computer Science, Artificial Intelligence

Deep U-Net architecture with curriculum learning for myocardial pathology segmentation in multi-sequence cardiac magnetic resonance images

Hengfei Cui et al.

Summary: This study proposes a deep learning-based segmentation method for myocardial pathology, which can automatically delineate the areas of myocardial infarction and edema. The method achieves good performance in myocardial pathology segmentation.

KNOWLEDGE-BASED SYSTEMS (2022)

Add to Collection

Article Computer Science, Information Systems

Spatio-Temporal Multi-Task Learning for Cardiac MRI Left Ventricle Quantification

Sulaiman Vesal et al.

Summary: The study presents a deep learning model to quantify cardiac LV morphology and detect cardiac phase cycle using spatio-temporal convolutions. The proposed method demonstrates high prediction accuracy, despite variations in cardiac morphology, image appearance, and contrast in cardiac MR sequences.

IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS (2021)

Add to Collection

Article Computer Science, Information Systems

DBAN: Adversarial Network With Multi-Scale Features for Cardiac MRI Segmentation

Xinyu Yang et al.

Summary: A novel deep adversarial network, DBAN, is proposed for left ventricle, right ventricle, and myocardium segmentation in short-axis cardiac MRI. Experimental results show that DBAN achieves state-of-the-art performance on the ACDC dataset, with cardiac function indices similar to those from clinical experts.

IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS (2021)

Add to Collection

Article Multidisciplinary Sciences