ISA-Net: Improved spatial attention network for PET-CT tumor segmentation

Huang, Zhengyong; Zou, Shurong; Wang, Guoshuai; Chen, Zixiang; Shen, Hao; Wang, Haiyan; Zhang, Na; Zhang, Lu; Yang, Fan; Wang, Haining; Liang, Dong; Niu, Tianye; Zhu, Xiaohua; Hu, Zhanli

doi:10.1016/j.cmpb.2022.107129

Cited by 15 publications

(7 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Finally, data volume augmentation is used to improve the model’s robustness and avoid overfitting. Data augmentation methods including flip, rotations, random noise, scaling, elastic deformations, and the mixed data (mix-up) method ( 46 , 47 ) are commonly used in PET/CT tumor segmentation tasks. The mix-up ( 48 ) technique can increase data diversity, which generates randomly weighted combined image pairs based on training data pairs.…”

Section: Discussionmentioning

confidence: 99%

Multi-modal tumor segmentation methods based on deep learning: a narrative review

Xue,

Yao,

Teng

2024

Quant Imaging Med Surg

View full text Add to dashboard Cite

Background and Objective Automatic tumor segmentation is a critical component in clinical diagnosis and treatment. Although single-modal imaging provides useful information, multi-modal imaging provides a more comprehensive understanding of the tumor. Multi-modal tumor segmentation has been an essential topic in medical image processing. With the remarkable performance of deep learning (DL) methods in medical image analysis, multi-modal tumor segmentation based on DL has attracted significant attention. This study aimed to provide an overview of recent DL-based multi-modal tumor segmentation methods. Methods In in the PubMed and Google Scholar databases, the keywords “multi-modal”, “deep learning”, and “tumor segmentation” were used to systematically search English articles in the past 5 years. The date range was from 1 January 2018 to 1 June 2023. A total of 78 English articles were reviewed. Key Content and Findings We introduce public datasets, evaluation methods, and multi-modal data processing. We also summarize common DL network structures, techniques, and multi-modal image fusion methods used in different tumor segmentation tasks. Finally, we conclude this study by presenting perspectives for future research. Conclusions In multi-modal tumor segmentation tasks, DL technique is a powerful method. With the fusion methods of different modal data, the DL framework can effectively use the characteristics of different modal data to improve the accuracy of tumor segmentation.

show abstract

Section: Discussionmentioning

confidence: 99%

Multi-modal tumor segmentation methods based on deep learning: a narrative review

Xue,

Yao,

Teng

2024

Quant Imaging Med Surg

View full text Add to dashboard Cite

show abstract

“…In this section, a series of ablation studies on the proposed YOLOS and extensive comparative experiments (including RetinaNet [24], Faster R-CNN [26], Cascade R-CNN [27], YOLOv3 [46], YOLOx [47], Mask-CNN [34], Cascade Mask-R-CNN [48], YOLACT [35], Seg-Former [32], ISA-Net [33], PID-Net [36], YOLOv5 [18], YOLOv7 [19], YOLOv8 [20] and YOLOv9 [21]) for station building surface defect detection are conducted to evaluate the performance of the YOLOS with a true railroad station images captured by UAV. Besides, the UAV imagery datasets acquisition, training details, and evaluation criteria are all presented in this section.…”

Section: Methodsmentioning

confidence: 99%

“…Furthermore, U 2 -Net [29], one of the many U-Net structure-based variants [30,31], achieves competitive effects in multiple segmented datasets through capturing more contextual information from different scales. To improve the semantic segmentation effect and reduce the amount of calculation, SegFormer [32] and ISANet [33] designed a lightweight multi-layer perceptron and Self-Attention mechanism respectively, which significantly increased the receptive field of the encoder. Besides, Mask R-CNN [34] that extends Faster R-CNN [27] to incorporate an additional branch for predicting segmentation masks, realizes high-precision instance-level segmentation, but only has an inference speed of 5 FPS.…”

Section: Related Workmentioning

confidence: 99%

UAV imagery-based railroad station building inspection using hybrid learning architecture

Liu,

Yang,

et al. 2024

Meas. Sci. Technol.

View full text Add to dashboard Cite

The apparent defects of railroad station building such as the broken glass, missing screw, and rust corrosion need to be inspected regularly and timely to ensure the safety of passenger. Unmanned Aerial Vehicle (UAV) imagery-based inspections have the potential to revolutionize current manual visual inspections by providing a better overhead view and mitigating safety concerns. This paper proposes a hybrid learning architecture called YOLOS (You Only Look Once Station Scene) to simultaneously detect and segment station building surface defects of UAV images. First, a novel Squeeze-and-Excitation (SE) attention block is integrated into the detection branch to adaptively learn the weights of the feature channels, thereby promoting the network to pay attention to the critical deep features of the objects. And then, a new semantic segmentation branch parallel to the detection branch is designed and assembled in YOLOS for pixel-level defect recognition. Finally, extensive experiments on railroad station dataset established with drone imagery prove the effectiveness and robustness of the SE-based network on station surface defect detection. This method can quickly convert UAV imagery into useful information with a high detection rate.

show abstract

“…Drawing on the efficacy of self-attention, scholars have suggested incorporating spatial-wise attention alongside channel-wise attention within transformer blocks to capture interactions [29] , [30] , [31] . These dual attention schemes based on self-attention have demonstrated improved performance across medical image segmentation tasks [32] , [33] , [34] , [35] , [36] . Taking inspiration from the aforementioned research, we propose a novel transformer block that leverages spatial-wise and channel-wise attention mechanisms to enhance the capture of both spatial and channel information in 3D medical images.…”

Section: Introductionmentioning

confidence: 99%

SCANeXt: Enhancing 3D medical image segmentation with dual attention network and depth-wise convolution

Liu,

Zhang,

Yue

et al. 2024

Heliyon

View full text Add to dashboard Cite

ISA-Net: Improved spatial attention network for PET-CT tumor segmentation

Cited by 15 publications

References 15 publications

Multi-modal tumor segmentation methods based on deep learning: a narrative review

Multi-modal tumor segmentation methods based on deep learning: a narrative review

UAV imagery-based railroad station building inspection using hybrid learning architecture

SCANeXt: Enhancing 3D medical image segmentation with dual attention network and depth-wise convolution

Contact Info

Product

Resources

About