Thyroid region prior guided attention for ultrasound segmentation of thyroid nodules

Article Computer Science, Interdisciplinary Applications

TypeSeg: A type-aware encoder-decoder network for multi-type ultrasound images co-segmentation

Fang Chen et al.

Summary: This study proposes a novel type-aware encoder-decoder network (TypeSeg) for multi-type ultrasound images co-segmentation, achieving excellent segmentation results through type-aware metric learning module and decision module.

COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE (2022)

添加到收藏夹

Article Computer Science, Interdisciplinary Applications

TNSNet: Thyroid nodule segmentation in ultrasound imaging using soft shape supervision

Jiawei Sun et al.

Summary: This study proposes a dual-path convolutional neural network for accurate segmentation of thyroid nodules on ultrasound images. By incorporating soft shape supervision blocks between the region and shape paths, the network can accurately detect and segment nodules by enhancing shape features and adding them as auxiliary information. The proposed network outperforms classical deep learning-based networks in terms of segmentation performance.

COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE (2022)

添加到收藏夹

Article Biology

Adaptive soft erasure with edge self-attention for weakly supervised semantic segmentation: Thyroid ultrasound image case study

Mei Yu et al.

Summary: The article introduces a weakly supervised segmentation neural network approach that addresses segmentation issues in medical images using a dual branch soft erase module and scale feature adaptation module, while further enhancing the nodule edge segmentation effect with an edge-based attention mechanism.

COMPUTERS IN BIOLOGY AND MEDICINE (2022)

添加到收藏夹

Article Biology

CSM-Net: Automatic joint segmentation of intima-media complex and lumen in carotid artery ultrasound images

Yanchao Yuan et al.

Summary: In this paper, the authors propose a flexible method CSM-Net for the joint segmentation of intima-media complex (IMC) and lumen in carotid ultrasound images. The method combines cascaded dilated convolutions, squeeze-excitation module, and triple spatial attention module to enhance feature extraction and improve segmentation accuracy. The proposed method achieves superior performance compared to state-of-the-art methods, demonstrating its potential utility in clinical IMC segmentation.

COMPUTERS IN BIOLOGY AND MEDICINE (2022)

添加到收藏夹

Article Acoustics

Tumor Microenvironment, Radiology, and Artificial Intelligence Should We Consider Tumor Periphery?

Afshin Mohammadi et al.

Summary: This study investigates the diagnostic potential of the tumor microenvironment (TME) in differentiating benign and malignant lesions using image quantification and machine learning. The results suggest that radiomic features extracted from extended contours outperform radiologists' contours in diagnosing breast and lymph node lesions. The study highlights the importance of the tumor periphery and TME in medical imaging diagnosis.

JOURNAL OF ULTRASOUND IN MEDICINE (2022)

添加到收藏夹

Article Automation & Control Systems

Integrate domain knowledge in training multi-task cascade deep learning model for benign-malignant thyroid nodule classification on ultrasound images

Wenkai Yang et al.

Summary: The study proposes a multi-task cascade deep learning model that integrates radiologists' domain knowledge and uses multimodal ultrasound images for automatic thyroid nodule diagnosis. Experimental results show that the model can achieve similar classification performance to fully supervised learning with only about 35% of labeled dataset, saving time and effort compared to traditional methods.

ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE (2021)

添加到收藏夹

Review Computer Science, Interdisciplinary Applications