Cascade convolutional neural networks for automatic detection of thyroid nodules in ultrasound images

Purpose It is very important for calculation of clinical indices and diagnosis to detect thyroid nodules from ultrasound images. However, this task is a challenge mainly due to heterogeneous thyroid nodules with distinct components are similar to background in ultrasound images. In this study, we employ cascade deep convolutional neural networks (CNNs) to develop and evaluate a fully automatic detection of thyroid nodules from 2D ultrasound images. Methods Our cascade CNNs are a type of hybrid model, consisting of two different CNNs and a new splitting method. Specifically, it employs a deep CNN to learn the segmentation probability maps from the ground true data. Then, all the segmentation probability maps are split into different connected regions by the splitting method. Finally, another deep CNN is used to automatically detect the thyroid nodules from ultrasound thyroid images. Results Experiment results illustrate the cascade CNNs are very effective in detection of thyroid nodules. Specially, the value of area under the curve of receiver operating characteristic is 98.51%. The Free‐response receiver operating characteristic (FROC) and jackknife alternative FROC (JAFROC) analyses show a significant improvement in the performance of our cascade CNNs compared to that of other methods. The multi‐view strategy can improve the performance of cascade CNNs. Moreover, our special splitting method can effectively separate different connected regions so that the second CNN can correctively gain the positive and negative samples according to the automatic labels. Conclusions The experiment results demonstrate the potential clinical applications of this proposed method. This technique can offer physicians an objective second opinion, and reduce their heavy workload so as to avoid misdiagnosis causes because of excessive fatigue. In addition, it is easy and reproducible for a person without medical expertise to diagnose thyroid nodules.

[1]  Georg Langs,et al.  Unsupervised Pre-training Across Image Domains Improves Lung Tissue Classification , 2014, MCV.

[2]  J. Strzelczyk.  The Essential Physics of Medical Imaging , 2003 .

[3]  Michalis A. Savelonas,et al.  A computer-aided system for malignancy risk assessment of nodules in thyroid US images based on boundary features , 2009, Comput. Methods Programs Biomed..

[4]  Nitish Srivastava,et al.  Improving neural networks by preventing co-adaptation of feature detectors , 2012, ArXiv.

[5]  Thomas J. Downey,et al.  Using the receiver operating characteristic to asses the performance of neural classifiers , 1999, IJCNN'99. International Joint Conference on Neural Networks. Proceedings (Cat. No.99CH36339).

[6]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[7]  Nikos Dimitropoulos,et al.  Variable Background Active Contour Model for Computer-Aided Delineation of Nodules in Thyroid Ultrasound Images , 2007, IEEE Transactions on Information Technology in Biomedicine.

[8]  Rob Fergus,et al.  Visualizing and Understanding Convolutional Networks , 2013, ECCV.

[9]  Savita Gupta,et al.  Automated delineation of thyroid nodules in ultrasound images using spatial neutrosophic clustering and level set , 2016, Appl. Soft Comput..

[10]  John F. Hamilton,et al.  A Free Response Approach To The Measurement And Characterization Of Radiographic Observer Performance , 1977, Other Conferences.

[11]  Savita Gupta,et al.  Computer-Aided Diagnosis of Thyroid Nodule: A Review , 2012 .

[12]  Shuiwang Ji,et al.  Deep convolutional neural networks for multi-modality isointense infant brain image segmentation , 2015, NeuroImage.

[13]  Luca Maria Gambardella,et al.  Mitosis Detection in Breast Cancer Histology Images with Deep Neural Networks , 2013, MICCAI.

[14]  Andrew Zisserman,et al.  Very Deep Convolutional Networks for Large-Scale Image Recognition , 2014, ICLR.

[15]  Gengsheng Qin,et al.  continuous-scale diagnostic test Comparison of non-parametric confidence intervals for the area under the ROC curve of a , 2010 .

[16]  Michal Strzelecki,et al.  Texture Analysis Methods - A Review , 1998 .

[17]  Dimitrios K. Iakovidis,et al.  Efficient and Effective Ultrasound Image Analysis Scheme for Thyroid Nodule Detection , 2007, ICIAR.

[18]  Jian Sun,et al.  Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[19]  Yann LeCun,et al.  Regularization of Neural Networks using DropConnect , 2013, ICML.

[20]  F. S. Cohen,et al.  Classification of Rotated and Scaled Textured Images Using Gaussian Markov Random Field Models , 1991, IEEE Trans. Pattern Anal. Mach. Intell..

[21]  David J. Hand,et al.  Measuring classifier performance: a coherent alternative to the area under the ROC curve , 2009, Machine Learning.

[22]  Andriy I. Bandos,et al.  On comparing methods for discriminating between actually negative and actually positive subjects with FROC type data. , 2008, Medical physics.

[23]  Fa Wu,et al.  Flip-Rotate-Pooling Convolution and Split Dropout on Convolution Neural Networks for Image Classification , 2015, ArXiv.

[24]  David Mackay,et al.  Probable networks and plausible predictions - a review of practical Bayesian methods for supervised neural networks , 1995 .

[25]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[26]  Yoshua Bengio,et al.  Gradient-based learning applied to document recognition , 1998, Proc. IEEE.

[27]  Nikos Dimitropoulos,et al.  Computer-aided thyroid nodule detection in ultrasound images , 2005, 18th IEEE Symposium on Computer-Based Medical Systems (CBMS'05).

[28]  Chuan-Yu Chang,et al.  Application of support-vector-machine-based method for feature selection and classification of thyroid nodules in ultrasound images , 2010, Pattern Recognit..

[29]  Kazutoshi Okamura,et al.  Quantitative analyses of sonographic images of the parotid gland in patients with Sjögren's syndrome. , 2006, Ultrasound in medicine & biology.

[30]  Jürgen Schmidhuber,et al.  Multi-column deep neural networks for image classification , 2012, 2012 IEEE Conference on Computer Vision and Pattern Recognition.

[31]  Jae Hyoung Kim,et al.  Differentiating Benign From Malignant Thyroid Nodules , 2012, Journal of ultrasound in medicine : official journal of the American Institute of Ultrasound in Medicine.

[32]  Dimitrios K. Iakovidis,et al.  Fusion of fuzzy statistical distributions for classification of thyroid ultrasound patterns , 2010, Artif. Intell. Medicine.

[33]  J. Sipos,et al.  Advances in ultrasound for the diagnosis and management of thyroid cancer. , 2009, Thyroid : official journal of the American Thyroid Association.

[34]  Διονύσης Α. Κάβουρας,et al.  Morphological and wavelet features towards sonographic thyroid nodules evaluation , 2015 .

[35]  Dimitrios K. Iakovidis,et al.  ΤND: A Thyroid Nodule Detection System for Analysis of Ultrasound Images and Videos , 2012, Journal of Medical Systems.

[36]  Dev P Chakraborty,et al.  Validation and statistical power comparison of methods for analyzing free-response observer performance studies. , 2008, Academic radiology.

[37]  David Gur,et al.  Area under the Free‐Response ROC Curve (FROC) and a Related Summary Index , 2009, Biometrics.

[38]  Wei Shen,et al.  Multi-scale Convolutional Neural Networks for Lung Nodule Classification , 2015, IPMI.

[39]  Anjan Biswas,et al.  Thyroid Nodule Recognition Based on Feature Selection and Pixel Classification Methods , 2013, Journal of Digital Imaging.

[40]  T. Teknos,et al.  Evaluation of the thyroid nodule. , 2006, Cancer control : journal of the Moffitt Cancer Center.

[41]  Z. Hall.  Cancer , 1906, The Hospital.

[42]  Luca Maria Gambardella,et al.  Deep Neural Networks Segment Neuronal Membranes in Electron Microscopy Images , 2012, NIPS.

[43]  Dev P Chakraborty,et al.  Observer studies involving detection and localization: modeling, analysis, and validation. , 2004, Medical physics.

[44]  Elli Angelopoulou,et al.  Using Power Watersheds to Segment Benign Thyroid Nodules in Ultrasound Image Data , 2011, Bildverarbeitung für die Medizin.

[45]  U Rajendra Acharya,et al.  Non-invasive automated 3D thyroid lesion classification in ultrasound: a class of ThyroScan™ systems. , 2012, Ultrasonics.