Shurui Bai
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Shurui Bai has a total of 14 co-authors. Here are the top co-authors by number of shared publications:
- Lan Ma
Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen, China
2 shared publications - Weihao Gao
Tsinghua University Shenzhen International Graduate School 518055 Guangzhou Shenzhen China
2 shared publications - Chucheng Chen
Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen 518055, China
1 shared publication - Fang Li
Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen 518055, China
1 shared publication - Fang Li
Institute of Biopharmaceutical and Health Engineering, Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen 518055, China
1 shared publication - Fuju Rong
Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen 518055, China
1 shared publication
Latest Publications
- Open Access
- 22 September 2024
- Published by MDPI AG in Bioengineering
- Vol. 11 (9) , 950
FTSNet: Fundus Tumor Segmentation Network on Multiple Scales Guided by Classification Results and Prompts
The segmentation of fundus tumors is critical for ophthalmic diagnosis and treatment, yet it presents unique challenges due to the variability in lesion size and shape. Our study introduces Fundus Tumor Segmentation Network (FTSNet), a novel segmentation network designed to address these challenges by leveraging classification results and prompt learning. Our key innovation is the multiscale feature extractor and the dynamic prompt head. Multiscale feature extractors are proficient in eliciting a spectrum of feature information from the original image across disparate scales. This proficiency is fundamental for deciphering the subtle details and patterns embedded in the image at multiple levels of granularity. Meanwhile, a dynamic prompt head is engineered to engender bespoke segmentation heads for each image, customizing the segmentation process to align with the distinctive attributes of the image under consideration. We also present the Fundus Tumor Segmentation (FTS) dataset, comprising 254 pairs of fundus images with tumor lesions and reference segmentations. Experiments demonstrate FTSNet’s superior performance over existing methods, achieving a mean Intersection over Union (mIoU) of 0.8254 and mean Dice (mDice) of 0.9042. The results highlight the potential of our approach in advancing the accuracy and efficiency of fundus tumor segmentation.
- 22 September 2024
- Published by MDPI AG in Bioengineering
- Vol. 11 (9) , 950
Open AccessFTSNet: Fundus Tumor Segmentation Network on Multiple Scales Guided by Classification Results and Prompts
The segmentation of fundus tumors is critical for ophthalmic diagnosis and treatment, yet it presents unique challenges due to the variability in lesion size and shape. Our study introduces Fundus Tumor Segmentation Network (FTSNet), a novel segmentation network designed to address these challenges by leveraging classification results and prompt learning. Our key innovation is the multiscale feature extractor and the dynamic prompt head. Multiscale feature extractors are proficient in eliciting a spectrum of feature information from the original image across disparate scales. This proficiency is fundamental for deciphering the subtle details and patterns embedded in the image at multiple levels of granularity. Meanwhile, a dynamic prompt head is engineered to engender bespoke segmentation heads for each image, customizing the segmentation process to align with the distinctive attributes of the image under consideration. We also present the Fundus Tumor Segmentation (FTS) dataset, comprising 254 pairs of fundus images with tumor lesions and reference segmentations. Experiments demonstrate FTSNet’s superior performance over existing methods, achieving a mean Intersection over Union (mIoU) of 0.8254 and mean Dice (mDice) of 0.9042. The results highlight the potential of our approach in advancing the accuracy and efficiency of fundus tumor segmentation.
- Open Access
FNeXter: A Multi-Scale Feature Fusion Network Based on ConvNeXt and Transformer for Retinal OCT Fluid Segmentation
The accurate segmentation and quantification of retinal fluid in Optical Coherence Tomography (OCT) images are crucial for the diagnosis and treatment of ophthalmic diseases such as age-related macular degeneration. However, the accurate segmentation of retinal fluid is challenging due to significant variations in the size, position, and shape of fluid, as well as their complex, curved boundaries. To address these challenges, we propose a novel multi-scale feature fusion attention network (FNeXter), based on ConvNeXt and Transformer, for OCT fluid segmentation. In FNeXter, we introduce a novel global multi-scale hybrid encoder module that integrates ConvNeXt, Transformer, and region-aware spatial attention. This module can capture long-range dependencies and non-local similarities while also focusing on local features. Moreover, this module possesses the spatial region-aware capabilities, enabling it to adaptively focus on the lesions regions. Additionally, we propose a novel self-adaptive multi-scale feature fusion attention module to enhance the skip connections between the encoder and the decoder. The inclusion of this module elevates the model’s capacity to learn global features and multi-scale contextual information effectively. Finally, we conduct comprehensive experiments to evaluate the performance of the proposed FNeXter. Experimental results demonstrate that our proposed approach outperforms other state-of-the-art methods in the task of fluid segmentation.
Open AccessFNeXter: A Multi-Scale Feature Fusion Network Based on ConvNeXt and Transformer for Retinal OCT Fluid Segmentation
The accurate segmentation and quantification of retinal fluid in Optical Coherence Tomography (OCT) images are crucial for the diagnosis and treatment of ophthalmic diseases such as age-related macular degeneration. However, the accurate segmentation of retinal fluid is challenging due to significant variations in the size, position, and shape of fluid, as well as their complex, curved boundaries. To address these challenges, we propose a novel multi-scale feature fusion attention network (FNeXter), based on ConvNeXt and Transformer, for OCT fluid segmentation. In FNeXter, we introduce a novel global multi-scale hybrid encoder module that integrates ConvNeXt, Transformer, and region-aware spatial attention. This module can capture long-range dependencies and non-local similarities while also focusing on local features. Moreover, this module possesses the spatial region-aware capabilities, enabling it to adaptively focus on the lesions regions. Additionally, we propose a novel self-adaptive multi-scale feature fusion attention module to enhance the skip connections between the encoder and the decoder. The inclusion of this module elevates the model’s capacity to learn global features and multi-scale contextual information effectively. Finally, we conduct comprehensive experiments to evaluate the performance of the proposed FNeXter. Experimental results demonstrate that our proposed approach outperforms other state-of-the-art methods in the task of fluid segmentation.