基于上下文感知的轻量化FAST腹腔积液与解剖结构联合识别

Context-aware lightweight FAST for joint recognition of ascites and anatomical structures

摘要

针对创伤重点超声评估（FAST）技术高度依赖操作者经验及现有深度学习模型多任务协同差、计算效率低、病灶识别定位能力不足等瓶颈，提出一种基于上下文感知的轻量化多任务模型（CALM-FAST），通过特征共享机制与任务特异性表征有机结合，实现腹腔积液区域分割与解剖结构检测的联合优化。研究在主干网络引入上下文感知轻量模块，结合快速卷积和特征融合策略，在保证全局特征表征能力的同时将计算复杂度降低至6.90 GFLOPs；提出多任务联合损失函数以缓解FAST数据不平衡问题；构建包含693张超声影像真实世界数据集，采用分层迁移学习策略增强模型泛化能力。结果表明，本文方法在解剖结构检测任务mAP@0.5和mAP@0.5:0.95分别达到93.0%和64.5%；腹腔积液分割任务IoU指标提升至25.9%，较基线模型提升9.0%。

Abstract

Focused assessment with sonography for trauma (FAST) technology highly depends on operator experience, and existing deep learning-based models encounter key technical challenges including inadequate multi-task collaboration, low computational efficiency, and insufficient ability to identify and locate lesions. To address the above issues, a context-aware lightweight multi-task model (CALM-FAST) is proposed, which combines feature sharing mechanism with task-specific representation to achieve joint optimization of peritoneal effusion segmentation and anatomical structure detection. The proposed model introduces a context-aware lightweight module into the backbone network, and integrates fast convolution and feature fusion strategies, which reduces the computational complexity to 6.90 GFLOPs while preserving global feature representation capability. Furthermore, a multi-task joint loss function is used to mitigate the data imbalance issue in FAST. In addition, a real-world dataset containing 693 ultrasound images is constructed, and a hierarchical transfer learning strategy is adopted to enhance the generalization performance. Results show that the proposed method achieves a mAP@0.5 of 93.0% and a mAP@0.5:0.95 of 64.5% in the anatomical structure detection task. For the peritoneal effusion segmentation, the IoU is increased to 25.9%, which is 9.0% higher than the baseline model.

关键词

创伤重点超声评估 / 腹部超声 / 上下文感知 / 多任务学习 / 轻量化

Key words

focused assessment with sonography for trauma / abdominal ultrasonography / context-aware / multi-task learning / lightweight

参考文献

[1] 李远建, 曹钰, 勾承锐 . 构架灾害事故的紧急医疗救援体系[J]. 中华急诊医学杂志, 2006, 15(6): 569-570.

[2] Li YJ, Cao Y, Gou CR . Constructing an emergency medical rescue system for disaster accidents[J]. Chinese Journal of Emergency Medicine, 2006, 15(6): 569-570.

[3] 王延明, 王能, 钱国军 . 便携式超声在战伤救治中应用[J]. 临床军医杂志, 2017, 45(1): 100-103.

[4] Wang YM, Wang N, Qian GJ . Application of portable ultrasound in war injury treatment[J]. Clinical Journal of Medical Officers, 2017, 45(1): 100-103.

[5] Cheng CT, Ooyang CH, Liao CH, et al. Applications of deep learning in trauma radiology: a narrative review[J]. Biomed J, 2025, 48(1): 100743.

[6] Helling TS, Wilson J, Augustosky K . The utility of focused abdominal ultrasound in blunt abdominal trauma: a reappraisal[J]. Am J Surg, 2007, 194(6): 728-732.

[7] Taye M, Morrow D, Cull J, et al. Deep learning for FAST quality assessment[J]. J Ultrasound Med, 2023, 42(1): 71-79.

[8] Gleeson T, Blehar D . Point—of—care ultrasound in trauma[J]. Semin Ultrasound CT MR, 2018, 39(4): 374-383.

[9] Cheng CY, Chiu IM, Hsu MY, et al. Deep learning assisted detection of abdominal free fluid in Morison's pouch during focused assessment with sonography in trauma[J]. Front Med (Lausanne), 2021, 8: 707437.

[10] Leo MM, Potter IY, Zahiri M, et al. Using deep learning to detect the presence and location of hemoperitoneum on the focused assessment with sonography in trauma (FAST) examination in adults[J]. J Digit Imaging, 2023, 36(5): 2035-2050.

[11] Yıldız Potter İ, Leo MM, Vaziri A, et al. Automated detection and localization of pericardial effusion from point—of—care cardiac ultrasound examination[J]. Med Biol Eng Comput, 2023, 61(8): 1947-1959.

[12] Lin ZY, Li ZY, Cao P, et al. Deep learning for emergency ascites diagnosis using ultrasonography images[J]. J Appl Clin Med Phys, 2022, 23(7): e13695.

[13] Chiu IM, Lin CH, Yau FF, et al. Use of a deep—learning algorithm to guide novices in performing focused assessment with sonography in trauma[J]. JAMA Netw Open, 2023, 6(3): e235102.

[14] Fu J, Liu J, Tian HJ, et al. Dual attention network for scene segmentation[C]// 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway, NJ, USA: IEEE, 2019: 3141-3149.

[15] Oktay O, Schlemper J, Folgoc LL, et al. Attention U—Net: learning where to look for the pancreas[EB/OL]. (2018—05—20). https://arxiv.org/abs/1804.03999.

[16] Azad R, Heidari M, Wu YL, et al. Contextual attention network: transformer meets U—Net[C]// Machine Learning in Medical Imaging. Cham: Springer Nature Switzerland, 2022: 377-386.

[17] Devlin J, Chang MW, Lee K, et al. Bert: Pre—training of deep bidirectional transformers for language understanding[C]//In Proceedings of the 2019 conference of the North American chapter of the association for computational linguistics: human language technologies. Minneapolis, Minnesota, USA: Association for Computational Linguistics, 2019: 4171-4186.

[18] Wang XL, Girshick R, Gupta A, et al. Non—local neural networks[C]// 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. IEEE, 2018: 7794-7803.

[19] Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need[C]// Proceedings of the 31st International Conference on Neural Information Processing Systems. Red Hook, NY, USA: Curran Associates Inc., 2017: 6000-6010.

[20] 于学忠, 郭树彬, 张波, 等 . 急诊超声标准操作规范[J]. 中国急救医学, 2013, 33(7): 577-591.

[21] Yu XZ, Guo SB, Zhang B, et al. Standard operating procedures for emergency ultrasound[J]. Chinese Journal of Critical Care Medicine, 2013, 33(7): 577-591.

[22] Chen JR, Kao SH, He H, et al. Run, don't walk: chasing higher FLOPS for faster neural networks[C]// 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway, NJ, USA: IEEE, 2023: 12021-12031.

[23] Lin TY, Dollár P, Girshick R, et al. Feature pyramid networks for object detection[C]// 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway, NJ, USA: IEEE, 2017: 936-944.

[24] Liu S, Qi L, Qin HF, et al. Path aggregation network for instance segmentation[C]// 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway, NJ, USA: IEEE, 2018: 8759-8768.

[25] Varghese R. YOLOv8: a novel object detection algorithm with enhanced performance and robustness[C]// 2024 International Conference on Advances in Data Engineering and Intelligent Computing Systems (ADICS). Piscataway, NJ, USA: IEEE, 2024: 1-6.

[26] Kiliçarslan S, Celik M . RSigELU: a nonlinear activation function for deep neural networks[J]. Expert Syst Appl, 2021, 174: 114805.

[27] Taud H, Mas JF. Multilayer perceptron (MLP)[M]//Camacho Olmedo MT, Paegelow M, Mas JF, et al. Geomatic Approaches for Modeling Land Change Scenarios. Cham: Springer International Publishing, 2018: 451-455.

[28] Lin TY, Goyal P, Girshick R, et al. Focal loss for dense object detection[C]// 2017 IEEE International Conference on Computer Vision (ICCV). Piscataway, NJ, USA: IEEE, 2017: 2999-3007.

[29] Salehi SS, Erdogmus D, Gholipour A . Tversky loss function for image segmentation using 3D fully convolutional deep networks[C]// Machine Learning in Medical Imaging. Cham: Springer International Publishing, 2017: 379-387.

[30] Mao A, Mohri M, Zhong YT . Cross—entropy loss functions: theoretical analysis and applications[C]// Proceedings of the 40th International Conference on Machine Learning. Chia Laguna Resort, Sardinia, Italy: PMLR, 2023: 23803-23828.

[31] Li X, Wang WH, Wu LJ, et al. Generalized focal loss: learning qualified and distributed bounding boxes for dense object detection[C]// Proceedings of the 34th International Conference on Neural Information Processing Systems. Red Hook, NY, USA: Curran Associates Inc., 2020: 21002-21012.

[32] Zheng ZH, Wang P, Liu W, et al. Distance—IoU loss: faster and better learning for bounding box regression[C]// Proceedings of the Thirty—Fourth AAAI Conference on Artificial Intelligence and the Thirty—Second Conference on Innovative Applications of Artificial Intelligence and the Tenth Symposium on Educational Advances in Artificial Intelligence. Palo Alto, CA, USA: AAAI Press, 2020: 12993-13000.

[33] Wang JY, Wu QM, Zhang N . You only look at once for real—time and generic multi—task[J]. IEEE Trans Veh Technol, 2024, 73(9): 12625-12637.

基金资助

重庆市科技创新重大研发项目(CSTB2025TIAD-STX0010)

国家自然科学基金(82502485)

陆军军医大学第二附属医院青年博士人才重点项目(2024YQB020)

重庆市沙坪坝区科学技术科技公关重点项目(20240105)