Application of deep learning to the diagnosis of cervical lymph node metastasis from thyroid cancer with CT
To develop a deep learning–based computer-aided diagnosis (CAD) system for use in the CT diagnosis of cervical lymph node metastasis (LNM) in patients with thyroid cancer.
A total of 995 axial CT images that included benign (n = 647) and malignant (n = 348) lymph nodes were collected from 202 patients with thyroid cancer who underwent CT for surgical planning between July 2017 and January 2018. The datasets were randomly split into training (79.0%), validation (10.5%), and test (10.5%) datasets. Eight deep convolutional neural network (CNN) models were used to classify the images into metastatic or benign lymph nodes. Pretrained networks were used on the ImageNet and the best-performing algorithm was selected. Class-specific discriminative regions were visualized with attention heatmap using a global average pooling method.
The area under the ROC curve (AUROC) for the tested algorithms ranged from 0.909 to 0.953. The sensitivity, specificity, and accuracy of the best-performing algorithm were all 90.4%, respectively. Attention heatmap highlighted important subregions for further clinical review.
A deep learning–based CAD system could accurately classify cervical LNM in patients with thyroid cancer on preoperative CT with an AUROC of 0.953. Whether this approach has clinical utility will require evaluation in a clinical setting.
• A deep learning–based CAD system could accurately classify cervical lymph node metastasis. The AUROC for the eight tested algorithms ranged from 0.909 to 0.953.
• Of the eight models, the ResNet50 algorithm was the best-performing model for the validation dataset with 0.953 AUROC. The sensitivity, specificity, and accuracy of the ResNet50 model were all 90.4%, respectively, in the test dataset.
• Based on its high accuracy of 90.4%, we consider that this model may be useful in a clinical setting to detect LNM on preoperative CT in patients with thyroid cancer.
KeywordsArtificial intelligence Lymphatic metastasis Thyroid cancer Multidetector computed tomography
Area under the receiver operating characteristic curve
Class activation mapping
Convolutional neural network
Fine needle aspiration
Lymph node metastasis
Papillary thyroid carcinoma
The authors state that this work was supported by the National Research Foundation of Korea (no. 2017R1C1B5016217).
Compliance with ethical standards
The scientific guarantor of this publication is Eun Ju Ha.
Conflict of interest
The authors of this manuscript declare no relationships with any companies, whose products or services may be related to the subject matter of the article.
Statistics and biometry
No complex statistical methods were necessary for this paper.
Written informed consent was obtained from all patients before they underwent US.
This study was approved by our institutional review board.
• case-control study
- 2.Li N, Du XL, Reitzel LR, Xu L, Sturgis EM (2013) Impact of enhanced detection on the increase in thyroid cancer incidence in the United States: review of incidence trends by socioeconomic status within the surveillance, epidemiology, and end results registry, 1980-2008. Thyroid 23:103–110CrossRefGoogle Scholar
- 3.Haugen BR, Alexander EK, Bible KC et al (2016) 2015 American Thyroid Association management guidelines for adult patients with thyroid nodules and differentiated thyroid Cancer: the American Thyroid Association guidelines task force on thyroid nodules and differentiated thyroid cancer. Thyroid 26:1–133CrossRefGoogle Scholar
- 9.Wong SC, Gatt A, Stamatescu V, McDonnell MD (2016) Understanding data augmentation for classification: when to warp? https://arXiv.org/abs/1609.08764. Accessed 19 Nov 2018
- 12.Lee Y, Kim JH, Baek JH et al (2018) Value of CT added to ultrasonography for the diagnosis of lymph node metastasis in patients with thyroid cancer. Head Neck. https://doi.org/10.1002/hed.25202
- 14.Miotto R, Wang F, Wang S, Jiang X, Dudley JT (2017) Deep learning for healthcare: review, opportunities and challenges. Brief BioinformGoogle Scholar
- 18.Jeong EY, Kim HL, Ha EJ et al (2018) Computer-aided diagnosis system for thyroid nodules on ultrasonography: diagnostic performance and reproducibility based on the experience level of operators. Eur RadiolGoogle Scholar
- 20.Simonyan K, Zisserman A (2014) Very deep convolutional networks for large-scale image recognition. https://arXiv.org/abs/1409.1556. Accessed 19 Nov 2018
- 21.He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognitionProceedings of the IEEE conference on computer vision and. Pattern Recogn:770–778Google Scholar
- 22.Chollet F (2017) Xception: deep learning with depthwise separable convolutions. https://arXiv.org/abs/1610.02357. Accessed 19 Nov 2018
- 23.Szegedy C, Vanhoucke V, Ioffe S, Shlens J, Wojna Z (2016) Rethinking the inception architecture for computer vision. https://arXiv.org/abs/1512.00567. Accessed 19 Nov 2018
- 24.Szegedy C, Ioffe S, Vanhoucke V, Alemi AA (2016) Inception-v4, inception-resnet and the impact of residual connections on learning. https://arXiv.org/abs/1602.07261. Accessed 19 Nov 2018
- 25.Iandola F, MoskewiczM,Karayev S, Girshick R, Darrell T, Keutzer K (2014) Densenet: Implementing efficient convnet descriptor pyramids. https://arXiv.org/abs/1404.1869. Accessed 19 Nov 2018
- 27.Abadi M, Barham P, Chen J et al (2016) Tensorflow: a system for large-scale machine learning. https://arXiv.org/abs/1605.08695. Accessed 19 Nov 2018