基于深度残差网络的皮肤癌黑色素瘤识别
|
摘要
作为皮肤癌黑色素瘤主要检查手段的皮肤镜图像存在显著性低、类内差异大和样本数据量少等问题,难以采用传统算法实现高准确的识别。深度学习算法引入皮肤癌症检测,提出了一种基于深度残差网络的黑色素瘤识别算法。该算法通过构建深度残差网络提取皮肤镜图像的高维特征,使用残差学习防止网络梯度退化、降低网络训练的难度,实现了黑色素瘤的有效识别。相关仿真实验结果表明:所提出的基于深度残差网络的黑色素瘤识别算法性能明显优于基于卷积神经网络传统的算法,具有更高的准确性、敏感性、特异性和鲁棒性。
The detection of skin cancer melanoma in dermoscopic images has the following problems: low significance, large intraclass differences, and less sample data. It is difficult to achieve high-accuracy melanoma recognition using traditional algorithms. In this paper, deep learning algorithm is introduced into skin cancer detection, and a melanoma identification algorithm based on deep residual network is proposed. The algorithm extracts high-dimensional features of dermoscopic images by constructing a deep residual network. Using residual learning can prevent network gradient degradation and reduces the difficulty of network training, so it can achieve effective identification of melanoma. The results of relevant simulation experiments show that the performance of melanoma recognition algorithm based on deep residual network presented in this paper is obviously superior to the traditional algorithm based on convolutional neural network. The proposed algorithm has higher accuracy, sensitivity, specificity and robustness in melanoma recognition.
The detection of skin cancer melanoma in dermoscopic images has the following problems: low significance, large intraclass differences, and less sample data. It is difficult to achieve high-accuracy melanoma recognition using traditional algorithms. In this paper, deep learning algorithm is introduced into skin cancer detection, and a melanoma identification algorithm based on deep residual network is proposed. The algorithm extracts high-dimensional features of dermoscopic images by constructing a deep residual network. Using residual learning can prevent network gradient degradation and reduces the difficulty of network training, so it can achieve effective identification of melanoma. The results of relevant simulation experiments show that the performance of melanoma recognition algorithm based on deep residual network presented in this paper is obviously superior to the traditional algorithm based on convolutional neural network. The proposed algorithm has higher accuracy, sensitivity, specificity and robustness in melanoma recognition.
引文
[1] ROGERS H W,WEINSTOCK M A,Feldman S R,et al.Incidence estimate of nonmelanoma skin cancer (keratinocyte carcinomas) in the US population,2012[J].JAMA dermatology,2015,151(10):1081-1086.
[2] JERANT A F,JOHNSON J T,DEMASTES S C,et al.Early detection and treatment of skin cancer[J].American family physician,2000,62(2) :357-368.
[3] SIEGEL R L,MILLER K D,JEMAL A.Cancer statistics,2016[J].CA:a cancer journal for clinicians,2016,66(1):7-30.
[4] BALCH C M,BUZAID A C,SOONG S J,et al.Final version of the American joint committee on cancer staging system for cutaneous melanoma[J].Journal of clinical oncology,2001,19(16):3635-3648.
[5] 汪晓妍,郑焕彰,黄晓洁,等.上下文窄带约束下腹主动脉瘤外轮廓分割[J].浙江工业大学学报,2015,43(6):642-648.
[6] 张剑华,张自然,汪晓妍,等.基于结构显著性的医学图像质量评价[J].浙江工业大学学报,2015,43(6):636-641.
[7] 龙胜春,傅佳琪,尧丽君.改进型K-Means算法在肠癌病理图像分割中的应用[J].浙江工业大学学报,2014,42(5):581-585.
[8] TOMMASI T,LA TORRE E,CAPUTO B.Melanoma recognition using representative and discriminative kernel classifiers[C]//REINHARD B,MILAN S.International Workshop on Computer Vision Approaches to Medical Image Analysis.Graz:Springer,2006:1-12.
[9] 胡言广.黑色素瘤图像特征提取算法研究[D].成都:电子科技大学,2015.
[10] SCHAEFER G,KRAWCZYK B,CELEBI M E,et al.An ensemble classification approach for melanoma diagnosis[J].Memetic computing,2014,6(4):233-240.
[11] ANTHIMOPOULOS M,CHRISTODOULIDIS S,EBNER L,et al.Lung pattern classification for interstitial lung diseases using a deep convolutional neural network[J].IEEE transactions on medical imaging,2016,35(5):1207-1216.
[12] DOU Q,CHEN H,YU L,et al.Automatic detection of cerebral microbleeds from MR images via 3D convolutional neural networks[J].IEEE transactions on medical imaging,2016,35(5):1182-1195.
[13] CHEN H,QI X,YU L,et al.Dcan:deep contour-aware networks for accurate gland segmentation[C]//Proceedings of the IEEE conference on Computer Vision and Pattern Recognition.Las Vegas:IEEE Computer Society,2016:2487-2496.
[14] CODELLA N,CAI J,ABEDINI M,et al.Deep learning,sparse coding,and SVM for melanoma recognition in dermoscopy images[C]//Proceedings of International Workshop on Machine Learning in Medical Imaging.Munich:Springer,2015:118-126.
[15] KRIZHEVSKY A,SUTSKEVER I,HINTON G E.Imagenet classification with deep convolutional neural networks[C]//Advances in neural information processing systems.Nevada:Neural Information Processing Systems,2012:1097-1105.
[16] KAWAHARA J,BENTAIEB A,HAMARNEH G.Deep features to classify skin lesions[C]//Proceedings of IEEE 13th International Symposium on Biomedical Imaging.Prague:IEEE,2016:1397-1400.
[17] GAL Y,ISLAM R,GHAHRAMANI Z.Deep bayesian active learning with image data[C]//PRECUP D,TEH Y W.Proceedings of 34th International Conference on Machine Learning.Sydney:PMLR,2017:1183-1192.
[18] DENG J,DONG W,SOCHER R,et al.Imagenet:a large-scale hierarchical image database[C]//Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.Miami:IEEE Computer Society,2009:248-255.
[19] SIMONYAN K,ZISSERMAN A.Very deep convolutional networks for large-scale image recognition[EB/OL].[2018-03-10].https://arxiv.org/pdf/1409.1556.pdf.
[20] LECUN Y,BOTTOU L,BENGIO Y,et al.Gradient-based learning applied to document recognition[J].Proceedings of the IEEE,1998,86(11):2278-2324.
[21] SZEGEDY C,LIU W,JIA Y,et al.Going deeper with convolutions[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Boston:IEEE Computer Society,2015:1-9.
[22] GLOROT X,BENGIO Y.Understanding the difficulty of training deep feedforward neural networks[C]//Proceedings of the 13th International Conference on Artificial Intelligence and Statistics.Chia Laguna Resort:JMLR,2010:249-256.
[23] HE K,ZHANG X,REN S,et al.Deep residual learning for image recognition[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Las Vegas:IEEE Computer Society,2016:770-778.
[24] VATANEN T,RAIKO T,VALPOLA H,et al.Pushing stochastic gradient towards second-order methods-backpropagation learning with transformations in nonlinearities[C]//Proceedings of International Conference on Neural Information Processing.Berlin:Springer,2013:442-449.
[25] IOFFE S,SZEGEDY C.Batch normalization:Accelerating deep network training by reducing internal covariate shift[C]//Proceedings of the 32nd International Conference on Machine Learning.Lille:JMLR,2015:448-456.
[26] GLOROT X,BORDES A,BENGIO Y.Deep sparse rectifier neural networks[C]//Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics.Fort Lauderdale:JMLR,2011:315-323.
[27] GUTMAN D,CODELLA N C F,CELEBI E,et al.Skin lesion analysis toward melanoma detection:a challenge at the international symposium on biomedical imaging (ISBI) 2016,hosted by the international skin imaging collaboration (ISIC)[EB/OL].[2018-03-10].https://arxiv.org/pdf/1605.01397.pdf.
[2] JERANT A F,JOHNSON J T,DEMASTES S C,et al.Early detection and treatment of skin cancer[J].American family physician,2000,62(2) :357-368.
[3] SIEGEL R L,MILLER K D,JEMAL A.Cancer statistics,2016[J].CA:a cancer journal for clinicians,2016,66(1):7-30.
[4] BALCH C M,BUZAID A C,SOONG S J,et al.Final version of the American joint committee on cancer staging system for cutaneous melanoma[J].Journal of clinical oncology,2001,19(16):3635-3648.
[5] 汪晓妍,郑焕彰,黄晓洁,等.上下文窄带约束下腹主动脉瘤外轮廓分割[J].浙江工业大学学报,2015,43(6):642-648.
[6] 张剑华,张自然,汪晓妍,等.基于结构显著性的医学图像质量评价[J].浙江工业大学学报,2015,43(6):636-641.
[7] 龙胜春,傅佳琪,尧丽君.改进型K-Means算法在肠癌病理图像分割中的应用[J].浙江工业大学学报,2014,42(5):581-585.
[8] TOMMASI T,LA TORRE E,CAPUTO B.Melanoma recognition using representative and discriminative kernel classifiers[C]//REINHARD B,MILAN S.International Workshop on Computer Vision Approaches to Medical Image Analysis.Graz:Springer,2006:1-12.
[9] 胡言广.黑色素瘤图像特征提取算法研究[D].成都:电子科技大学,2015.
[10] SCHAEFER G,KRAWCZYK B,CELEBI M E,et al.An ensemble classification approach for melanoma diagnosis[J].Memetic computing,2014,6(4):233-240.
[11] ANTHIMOPOULOS M,CHRISTODOULIDIS S,EBNER L,et al.Lung pattern classification for interstitial lung diseases using a deep convolutional neural network[J].IEEE transactions on medical imaging,2016,35(5):1207-1216.
[12] DOU Q,CHEN H,YU L,et al.Automatic detection of cerebral microbleeds from MR images via 3D convolutional neural networks[J].IEEE transactions on medical imaging,2016,35(5):1182-1195.
[13] CHEN H,QI X,YU L,et al.Dcan:deep contour-aware networks for accurate gland segmentation[C]//Proceedings of the IEEE conference on Computer Vision and Pattern Recognition.Las Vegas:IEEE Computer Society,2016:2487-2496.
[14] CODELLA N,CAI J,ABEDINI M,et al.Deep learning,sparse coding,and SVM for melanoma recognition in dermoscopy images[C]//Proceedings of International Workshop on Machine Learning in Medical Imaging.Munich:Springer,2015:118-126.
[15] KRIZHEVSKY A,SUTSKEVER I,HINTON G E.Imagenet classification with deep convolutional neural networks[C]//Advances in neural information processing systems.Nevada:Neural Information Processing Systems,2012:1097-1105.
[16] KAWAHARA J,BENTAIEB A,HAMARNEH G.Deep features to classify skin lesions[C]//Proceedings of IEEE 13th International Symposium on Biomedical Imaging.Prague:IEEE,2016:1397-1400.
[17] GAL Y,ISLAM R,GHAHRAMANI Z.Deep bayesian active learning with image data[C]//PRECUP D,TEH Y W.Proceedings of 34th International Conference on Machine Learning.Sydney:PMLR,2017:1183-1192.
[18] DENG J,DONG W,SOCHER R,et al.Imagenet:a large-scale hierarchical image database[C]//Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.Miami:IEEE Computer Society,2009:248-255.
[19] SIMONYAN K,ZISSERMAN A.Very deep convolutional networks for large-scale image recognition[EB/OL].[2018-03-10].https://arxiv.org/pdf/1409.1556.pdf.
[20] LECUN Y,BOTTOU L,BENGIO Y,et al.Gradient-based learning applied to document recognition[J].Proceedings of the IEEE,1998,86(11):2278-2324.
[21] SZEGEDY C,LIU W,JIA Y,et al.Going deeper with convolutions[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Boston:IEEE Computer Society,2015:1-9.
[22] GLOROT X,BENGIO Y.Understanding the difficulty of training deep feedforward neural networks[C]//Proceedings of the 13th International Conference on Artificial Intelligence and Statistics.Chia Laguna Resort:JMLR,2010:249-256.
[23] HE K,ZHANG X,REN S,et al.Deep residual learning for image recognition[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Las Vegas:IEEE Computer Society,2016:770-778.
[24] VATANEN T,RAIKO T,VALPOLA H,et al.Pushing stochastic gradient towards second-order methods-backpropagation learning with transformations in nonlinearities[C]//Proceedings of International Conference on Neural Information Processing.Berlin:Springer,2013:442-449.
[25] IOFFE S,SZEGEDY C.Batch normalization:Accelerating deep network training by reducing internal covariate shift[C]//Proceedings of the 32nd International Conference on Machine Learning.Lille:JMLR,2015:448-456.
[26] GLOROT X,BORDES A,BENGIO Y.Deep sparse rectifier neural networks[C]//Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics.Fort Lauderdale:JMLR,2011:315-323.
[27] GUTMAN D,CODELLA N C F,CELEBI E,et al.Skin lesion analysis toward melanoma detection:a challenge at the international symposium on biomedical imaging (ISBI) 2016,hosted by the international skin imaging collaboration (ISIC)[EB/OL].[2018-03-10].https://arxiv.org/pdf/1605.01397.pdf.