改进YOLOv7的结直肠息肉检测算法

doi:10.3778/j.issn.1002-8331.2308-0273

摘要/Abstract

摘要： 计算机辅助诊断对提高息肉诊断准确率和降低结直肠癌死亡率至关重要，但息肉形态各异，息肉类似物和肠内的复杂环境导致目前的方法存在较多的误诊和漏诊。因此提出了一种改进的YOLOv7结直肠息肉检测算法（YOLOv7-IDH），使用含隐式知识的高效解耦头，充分利用隐含信息并防止分类和回归任务之间相互干扰；引入全局注意力机制，增强模型对浅层特征的提取能力；对SPPCSPC模块进行优化，减少模型参数和提高收敛速度。实验结果表明，改进模型在组合数据集上的F1分数和mAP@0.5分别达到了94.8%和97.1%，可以满足息肉自动检测的要求。

关键词: 息肉检测, 深度学习, 计算机辅助诊断, 解耦头, 注意力机制

Abstract: Computer-aided diagnosis is essential to improve polyp diagnostic accuracy and reduce colorectal cancer mortality, but the variety of polyp morphologies, polyp analogs and the complex environment in the bowel lead to more misdiagnosis and underdiagnosis with current methods. Therefore, an improved YOLOv7 colorectal polyp detection algorithm (YOLOv7-IDH) is proposed, which firstly, efficient decoupled heads with implicit knowledge are used to make full use of the implicit information and to prevent mutual interference between classification and regression tasks; then, global attention mechanism is introduced to enhance the model’s capability of extracting shallow features; finally, the SPPCSPC module is optimized to reduce the model parameters and to improve the convergence speed. The experimental results show that the F1 score and mAP@0.5 of the improved model on the combined dataset reach 94.8% and 97.1%, respectively, which can meet the requirements for automatic polyp detection.

Key words: polyp detection, deep learning, computer-aided diagnosis, decoupled head, attention mechanism

薛钦原, 胡珊珊, 胡新军, 严松才. 改进YOLOv7的结直肠息肉检测算法[J]. 计算机工程与应用, 2025, 61(1): 243-251.

XUE Qinyuan, HU Shanshan, HU Xinjun, YAN Songcai. Improved YOLOv7 Algorithm for Colorectal Polyp Detection[J]. Computer Engineering and Applications, 2025, 61(1): 243-251.

参考文献

[1] 龙思丹, 孙希珍, 赵栋燕, 等. 肠息肉病因学相关性研究进展[J]. 医学综述, 2020, 26(14): 2728-2732.
LONG S D, SUN X Z, ZHAO D Y, et al. Research progress in etiological correlation of intestinal polyps[J]. Medical Recapitulate, 2020, 26(14): 2728-2732.
[2] FREEDMAN D, BLAU Y, KATZIR L, et al. Detecting deficient coverage in colonoscopies[J]. IEEE Transactions on Medical Imaging, 2020, 39(11): 3451-3462.
[3] JIANG W, XIN L, ZHU S, et al. Risk factors related to polyp miss rate of short-term repeated colonoscopy[J]. Digestive Diseases and Sciences, 2023, 68(5): 2040-2049.
[4] SHIN Y, QADIR H A, AABAKKEN L, et al. Automatic colon polyp detection using region based deep CNN and post learning approaches[J]. IEEE Access, 2018, 6: 40950-40962.
[5] LIU X, GUO X, LIU Y, et al. Consolidated domain adaptive detection and localization framework for cross-device colonoscopic images[J]. Medical Image Analysis, 2021, 71: 102052.
[6] QADIR H A, SHIN Y, SOLHUSVIK J, et al. Toward real-time polyp detection using fully CNNs for 2D Gaussian shapes prediction[J]. Medical Image Analysis, 2021, 68: 101897.
[7] PACAL I, KARABOGA D. A robust real-time deep learning based automatic polyp detection system[J]. Computers in Biology and Medicine, 2021, 134: 104519.
[8] RAHMAN M M, WADUD M A H, HASAN M M. Computerized classification of gastrointestinal polyps using stacking ensemble of convolutional neural network[J]. Informatics in Medicine Unlocked, 2021, 24: 100603.
[9] HACKING S M, CHAKRABORTY B, NASIM R, et al. A holistic appraisal of stromal differentiation in colorectal cancer: biology, histopathology, computation, and genomics[J]. Pathology-Research and Practice, 2021, 220: 153378.
[10] WANG C Y, BOCHKOVSKIY A, LIAO H Y M. YOLOv7: trainable bag-of-freebies sets new state-of-the-art for real-time object detectors[J]. arXiv:2207.02696, 2022.
[11] LIU Y, SHAO Z, HOFFMANN N. Global attention mechanism: retain information to enhance channel-spatial interactions[J]. arXiv:2112.05561, 2021.
[12] GAO P, LU J, LI H, et al. Container: context aggregation network[J]. arXiv:2106.01401, 2021.
[13] DOLLáR P, SINGH M, GIRSHICK R. Fast and accurate model scaling[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021.
[14] VASU P K A, GABRIEL J, ZHU J, et al. An improved one millisecond mobile backbone[J]. arXiv:2206.04040, 2022.
[15] SONG G, LIU Y, WANG X. Revisiting the sibling head in object detector[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020.
[16] WU Y, CHEN Y, YUAN L, et al. Rethinking classification and localization for object detection[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020.
[17] TIAN Z, SHEN C, CHEN H, et al. FCOS: fully convolutional one-stage object detection[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision, 2019.
[18] GE Z, LIU S, WANG F, et al. YOLOX: exceeding YOLO series in 2021[J]. arXiv:2107.08430, 2021.
[19] WANG C Y, YEH I H, LIAO H Y M. You only learn one representation: unified network for multiple tasks[J]. arXiv:2105.04206, 2021.
[20] NIU Z, ZHONG G, YU H. A review on the attention mechanism of deep learning[J]. Neurocomputing, 2021, 452: 48-62.
[21] WOO S, PARK J, LEE J Y, et al. CBAM: convolutional block attention module[C]//Proceedings of the European Conference on Computer Vision, 2018.
[22] HE K, ZHANG X, REN S, et al. Spatial pyramid pooling in deep convolutional networks for visual recognition[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 37(9): 1904-1916.
[23] WANG C Y, LIAO H Y M, WU Y H, et al. CSPNet: a new backbone that can enhance learning capability of CNN[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2020.
[24] SILVA J, HISTACE A, ROMAIN O, et al. Toward embedded detection of polyps in WCE images for early diagnosis of colorectal cancer[J]. International Journal of Computer Assisted Radiology and Surgery, 2014, 9: 283-293.
[25] BERNAL J, SáNCHEZ F J, FERNáNDEZ-ESPARRACH G, et al. WM-DOVA maps for accurate polyp highlighting in colonoscopy: validation vs. saliency maps from physicians[J]. Computerized Medical Imaging and Graphics, 2015, 43: 99-111.
[26] JHA D, SMEDSRUD P H, RIEGLER M A, et al. Kvasir-SEG: a segmented polyp dataset[C]//Proceedings of the 26th International Conference on Multimedia Modeling, Daejeon, South Korea, 2020.
[27] LIU W, ANGUELOV D, ERHAN D, et al. SSD: single shot multibox detector[C]//Proceedings of the 14th European Conference on Computer Vision, Amsterdam, The Netherlands, 2016.
[28] CARION N, MASSA F, SYNNAEVE G, et al. End-to-end object detection with transformers[C]//Proceedings of the European Conference on Computer Vision, 2020.
[29] LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection[C]//Proceedings of the IEEE International Conference on Computer Vision, 2017.
[30] GLENN J. YOLOv5 release v6.1[EB/OL].[2023-0510]. https://github.com/ultralytics/yolov5/releases/tag/v6.1.
[31] LI C, LI L, JIANG H, et al. YOLOv6: a single-stage object detection framework for industrial applications[J]. arXiv:2209.02976, 2022.
[32] JIA X, MAI X, CUI Y, et al. Automatic polyp recognition in colonoscopy images using deep learning and two-stage pyramidal feature prediction[J]. IEEE Transactions on Automation Science and Engineering, 2020, 17(3): 1570-1584.
[33] WANG D, ZHANG N, SUN X, et al. AFP-Net: realtime anchor-free polyp detection in colonoscopy[C]//Proceedings of the 2019 IEEE 31st International Conference on Tools with Artificial Intelligence, 2019.
[34] YANG K, CHANG S, TIAN Z, et al. Automatic polyp detection and segmentation using shuffle efficient channel attention network[J]. Alexandria Engineering Journal, 2022, 61(1): 917-926.