颞骨CT内面神经、迷路、听骨结构深度学习的自动化分割方法

柯嘉; 吕弈; 杜雅丽; 王君臣; 王江; 孙世龙; 马芙蓉

doi:10.16098/j.issn.0529-1356.2020.05.003

PDF(5389 KB)

解剖学报 ›› 2020, Vol. 51 ›› Issue (5) : 653-658. DOI: 10.16098/j.issn.0529-1356.2020.05.003

解剖学与耳鼻喉科学

颞骨CT内面神经、迷路、听骨结构深度学习的自动化分割方法

柯嘉¹ 吕弈² 杜雅丽¹ 王君臣² 王江¹ 孙世龙¹ 马芙蓉^1*

作者信息 +

Automatic segmentation of facial nerve, labyrinthine and ossicles in temporal CT by deep learning

KE Jia¹ Lü Yi² DU Ya-li¹ WANG Jun-chen² WANG Jiang¹ SUN Shi-long¹ MA Fu-rong^1*

Author information +

文章历史 +

摘要

目的探讨神经网络的深度学习方法，进行颞骨CT内面神经、迷路及听骨结构的自动化分割的可行性和精确性。方法选择常规颞骨CT检查患者的数据，随机分为两组，一组为训练集（20例），另一组为测试集（5例）。在上述颞骨CT中采用手工分割的方法，分割出迷路、听骨及面神经结构。选择三维卷积神经网络3D U-Net作为深度学习中的神经网络结构部分，通过对训练集的训练，得到该网络的平均精度。用该网络模型对5组测试集中的不同解剖标志自动分割的结果与手工分割的结果进行测试，分别获得面神经、迷路及听小骨的测试精度。并将上述精度与另一种基于三维卷积神经网络结构的V-Net网络模型获得的精度进行比较。结果在颞骨CT标本中，采用面神经、迷路及听小骨分别对3D U-Net-plus和V-Net网络结构的自动分割进行训练，在训练样本中，3D U-Net-plus网络结构的平均误差为0.016，V-Net网络结构的平均误差为0.035，两者差异有统计学意义（P<0.05）；利用3D U-Net-plus神经网络自动分割的迷路、听小骨及面神经与手工分割图像的Dice相似指数分别为0.618±0.107、0.584±0.089和0.313±0.069，利用V-Net神经网络自动分割的迷路、听小骨、面神经与手工分割图像的Dice相似指数分别为0.322±0.089、0.176±0.100和0.128± 0.077，两者差异有统计学意义（P<0.001）。结论采用3D U-Net-plus神经网络，在颞骨内听骨、迷路及面神经的自动识别和分割方面具有可行性，该方法优于V-Net神经网络。随着网络结构的优化和学习样本的扩大，其将更加接近人工分割的效果。

Abstract

Objective To study the effect of deep learning based on neural network on automatic segmentation of facial nerve, labyrinth and ossicles in temporal CT. Methods The data of patients with conventional temporal bone CT examination were randomly divided into two groups, one was the training set (20 cases) and the other was the test set (5 cases). The structures of labyrinth, ossicles and facial nerve were segmented manually. The convolutional neural network 3D U-Net was selected as the neural network structure part in deep learning, and the average accuracy of the network was obtained through the training of the training set. The result of automatic and manual segmentation of 3 above anatomical markers in 5 test sets were tested by two network model, and the accuracy of facial nerve, labyrinth and ossicles were obtained respectively. The accuracy was compared with that obtained by the other 3D convolutional neural network V-Net network model. Results In the temporal CT, facial nerve, labyrinth and ossicles were used to train the automatic segmentation of 3D U-Net-plus and V-Net network respectively. In the training samples, the mean error of 3D U-Net-plus network was 0.016, and 0.035 by V-Net network, the difference was significant, P<0.05. The Dice similarity coefficient of labyrinth, ossicles and facial nerve with manual segmented images by 3D U-Net-plus neural network were 0.618±0.107, 0.584±0.089 and 0.313±0.069, and 0.322±0.089, 0.176±0.100 and 0.128± 0.077 by V-Net neural network. The segmentation effect of 3D U-Net-plus neural network was significantly better than that by V-Net network, P<0.001. Conclusion Using 3D U-Net-plus neural network, the ossicles, labyrinth and facial nerves in the temporal CT can be recognized and automatics segmantation quickly and effectively. This method is better than V-Net neural network. and more close to manuall segmentation. With the optimization of network structure and the expansion of learning samples, it will be closer to the effect of manual segmentation.

导出引用

柯嘉吕弈杜雅丽王君臣王江孙世龙马芙蓉. 颞骨CT内面神经、迷路、听骨结构深度学习的自动化分割方法[J]. 解剖学报. 2020, 51(5): 653-658 https://doi.org/10.16098/j.issn.0529-1356.2020.05.003

KE Jia Lü Yi DU Ya-li WANG Jun-chen WANG Jiang SUN Shi-long MA Fu-rong. Automatic segmentation of facial nerve, labyrinthine and ossicles in temporal CT by deep learning[J]. Acta Anatomica Sinica. 2020, 51(5): 653-658 https://doi.org/10.16098/j.issn.0529-1356.2020.05.003

中图分类号： R445.3

参考文献

［1］ Liu Y, Chen Sh. Review of medical image segmentation method［J］. Electronic Science and Technology, 2017,(8):169-172. (in Chinese)
刘宇, 陈胜. 医学图像分割方法综述［J］.电子科技,2017,(8):169-172.
［2］ Pan YL, Lu Y. The application of computer aided diagnosis with artificial intelligence in medical imaging［J］. International Journal of Medical Radiology, 2019, 42(1):3-7. (in Chinese)

潘亚玲, 陆勇. 人工智能在医学影像CAD中的应用［J］.国际医学放射学杂志,2019,42(1):3-7.

［3］ Shelhamer E, Long J, Darrell T. Fully convolutional networks for semantic segmentation［J］. IEEE Trans Pattern Anal Mach Intell, 2017,39(4):640-651.

［4］ Ronneberger O, Fischer P, Brox T. U-Net: convolutional networks for biomedical image segmentation［C］. Medical Image Computing and Computer-Assisted Intervention (MICCAI), 2015, 9351: 234-241.
［5］ Rudra PK, Poudel, Pablo L, et al. Recurrent fully convolutional neural networks for multi-slice MRI cardiac segmentation［C］. RAMBO 2016, HVSMR 2016: Reconstruction, Segmentation, and Analysis of Medical Images, 2017 10129: 83-94.
［6］ Kingma DP, Ba J. Adam: a method for stochastic optimization［J］. 3rd International Conference for Learning Representations（ICLR）, San Diego, 2015: 1-15.
［7］ Litjens G, Kooi T, Bejnordi BE, et al. A survey on deep learning in medical image analysis［J］. Med Image Anal, 2017,42: 60-88.
［8］ Gong JCh, Zhao ShY, Wang YJ. Research progress on deep learning-based medical image segmentation［J］. Chinese Journal of Medical Physics, 2019,36(4):420-424. (in Chinese)
宫进昌, 赵尚义, 王远军. 基于深度学习的医学图像分割研究进展［J］.中国医学物理学杂志,2019,36(4):420-424.
［9］ Guo W, Ju ZhJ, Wu QN, et al. Research progress of automatic organ image segmentation based on deep learning［J］. Chinese Medical Equipment Journal, 2020,41(1):85-94. (in Chinese)
郭雯, 鞠忠建, 吴青南, 等. 基于深度学习的器官自动分割研究进展［J］.医疗卫生装备,2020,41(1):85-94.
［10］ Noble JH, Warren FM, Labadie RF, et al. Automatic segmentation of the facial nerve and chorda tympani in CT images using spatially dependent feature values［J］. Med Phys, 2008,35(12):5375-5384.
［11］ Reda FA, Noble JH, Rivas A, et al. Automatic segmentation of the facial nerve and chorda tympani in pediatric CT scans［J］. Med Phys, 2011,38(10):5590-5600.
［12］ Noble JH, Labadie RF, Majdani O, et al. Automatic segmentation of intra-cochlear anatomy in conventional CT［J］. IEEE Trans Biomed Eng, 2011, 58(9): 2625-2632.
［13］ Noble JH, Dawant BM, Warren FM, et al. Automatic identification and 3D rendering of temporal bone anatomy［J］. Otol Neurotol, 2009,30(4):436-442.
［14］ Powell KA, Liang T, Hittle B, et al. Atlas-based segmentation of temporal bone anatomy［J］. Int J Comput Assist Radiol Surg, 2017,12(11):1937-1944.
［15］ Milletari F, Navab N, Ahmadi S. V-Net: fully convolutional neural networks for volumetric medical image segmentation［C］. 2016 Fourth International Conference on 3D Vision (3DV), Stanford, CA, 2016: 565-571.
［16］ ?i?ek ?, Abdulkadir A, Soeren S, et al. 3D U-Net: Learning dense volumetric segmentation from sprse annotation［C］. Medical Image Computing and ComputerAssisted Intervention(MICCAI), 2016, 2016, 9901: 424-432.
［17］ Fauser J, Stenin Ⅰ, Bauer M, et al. Toward an automatic preoperative pipeline for image-guided temporal bone surgery［J］. Int J Comput Assist Radiol Surg, 2019,14(6):967-976.