体育赛事投注记录

advertisement

Skin Lesion Classification: A Transfer Learning Approach Using EfficientNets

  • Vandana MiglaniEmail author
  • MPS Bhatia
Conference paper
  • 61 Downloads
Part of the Advances in Intelligent Systems and Computing book series (AISC, volume 1141)

Abstract

This paper studies the ability of deep convolutional neural networks (DCNNs) to classify skin lesions belonging to seven different categories. Two pre-trained state-of-the-art architectures for computer vision tasks ResNet-50 and Google’s recently proposed, EfficientNet-B0, were fine-tuned for the classification task on the HAM10000 dataset. The dataset comprises 10015 dermatoscopic images belonging to seven classes of skin cancer melanocytic nevus, melanoma, benign keratosis, basal cell carcinoma, actinic keratosis, vascular lesions, and dermatofibroma. The aim of the study was to establish how well the EfficientNet family of models (which result in up to 8.4\(\times \) parameter reduction and 16\(\times \)体育赛事投注记录 FLOPS reduction) transfers to the skin classification task in comparison with the ResNet architecture. Overall, it was found that the EfficientNet-B0 model, with fewer parameters, outperformed the ResNet-50 model. EfficientNet-B0 model produced better ROC AUC values for each classification category and also achieved higher macro and micro averaged AUC values for the overall classification, 0.93 and 0.97, respectively (in comparison with, 0.91 and 0.96 of the ResNet-50 model).

Keywords

Skin cancer Transfer learning Deep learning Lesion classification EfficieNtnet ResNet 

References

  1. 1.
    Nami, N., Giannini, E., Burroni, M., Fimiani, M., Rubegni, P.: Teledermatology: state-of-the-art and future perspectives. Expert. Rev. Dermatol. 7(1), 1–3 (2014).  
  2. 2.
    Fabbrocini, G., Triassi, M., Mauriello, M.C., Torre, G., Annunziata, M.C., De Vita, V., Pastore, F., D’Arco, V., Monfrecola, G.: Epidemiology of skin cancer: role of some environmental factors. Cancers (Basel) 2(4), 1980–1989 (2010).  
  3. 3.
    Melanoma research gathers momentum. Lancet. 385(2323) (2015).  
  4. 4.
    Ouhib, Z., et al.: Aspects of dosimetry and clinical practice of skin brachytherapy: the American Brachytherapy society working group report. 14(6), 840–858 (2015)
  5. 5.
    Dorj, U.O., Lee, K.K., Choi, J.Y., Lee, M.J.M.T. and Applications.: The skin cancer classification using deep convolutional neural network. 1–16 (2018)
  6. 6.
    Esteva, A. et al., Dermatologist-level classification of skin cancer with deep neural networks. 542(7639), 115 (2017)
  7. 7.
    Ruiz, D., Berenguer, V., Soriano, A., SNchez, B.J.E.S.W.A.: A decision support system for the diagnosis of melanoma: a comparative approach. 38(12), 15217–15223 (2011)
  8. 8.
    Argenziano, G., Soyer, H.P.: Dermoscopy of pigmented skin lesions: a valuable tool for early diagnosis of melanoma. Lancet Oncol. 2(7), 443–449 (2001)
  9. 9.
    Ali, A.R.A., Deserno, T.M.: A systematic review of automated melanoma detection in dermatoscopic images and its ground truth data. Proc. SPIE. Int. Soc. Opt. Eng. 28(8318), 1–6 (2012).  
  10. 10.
    Fabbrocini, G., De Vita, V., Pastore, F., D’Arco, V., Mazzella, C., Annunziata, M.C., Cacciapuoti, S., Mauriello, M.C., Monfrecola, A.: Teledermatology: from prevention to diagnosis of nonmelanoma and melanoma skin cancer. Int. J. Telemed. Appl. 2011(17), 125762 (2011).  
  11. 11.
    He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 770–778 (2016)
  12. 12.
    Tan, M., Le, Q.V.: Efficientnet: rethinking model scaling for convolutional neural networks. arXiv preprint (2019)
  13. 13.
    Tschandl, P., Rosendahl, C. Kittler, H.: The HAM10000 dataset, a large collection of multi-source dermatoscopic images of common pigmented skin lesions. Sci. Data 5(180161), (2018).  
  14. 14.
    Almansour, E., Jaffar, M.A.J.I.I.J.C.S.N.S.: Classification of dermoscopic skin cancer images using color and hybrid texture features. 16(4), 135–139 (2016)
  15. 15.
    Abbas, Q., Celebi, M.E., Serrano, C., GarcA, I.F., Ma, G.J.P.R.: Pattern classification of dermoscopy images: a perceptually uniform model. 46(1), 86–97 (2013)
  16. 16.
    Krizhevsky, A., Sutskever, I., Hinton, G.E.: Imagenet classification with deep convolutional neural networks. In: Advances in neural information processing systems, pp. 1097–1105 (2012)
  17. 17.
    Yu, L., Chen, H., Dou, Q., Qin, J., Heng, P.A.: Automated melanoma recognition in dermoscopy images via very deep residual networks. 36(4), 994–1004 (2017)
  18. 18.
    Haenssle, H., et al.: Man against machine: diagnostic performance of a deep learning convolutional neural network for dermoscopic melanoma recognition in comparison to 58 dermatologists. 29(8), 1836–1842 (2018)
  19. 19.
    Dorj, U.O., Lee, K.K., Choi, J.Y., Lee, M.J.M.T.: The skin cancer classification using deep convolutional neural network. 1–16 (2018)
  20. 20.
    Han, S.S., Kim, M.S., Lim, W., Park, G.H., Park, I., Chang, S.E.J.J.o.I.D.: Classification of the clinical images for benign and malignant cutaneous tumors using a deep learning algorithm. (2018)
  21. 21.
    Skin Cancer Homepage. . Accessed 10 Aug 2019
  22. 22.
    Skin Cancer Facts and Statisctis. . Last accessed 10 Aug 2019
  23. 23.
    Rezvantalab, A., Safigholi, H., Karimijeshni, S.: Dermatologist level dermoscopy skin cancer classification using different deep learning convolutional neural networks algorithms. (2018)

Copyright information

© Springer Nature Singapore Pte Ltd. 2021

Authors and Affiliations

  1. 1.Netaji Subhas University of TechnologyNew DelhiIndia

Personalised recommendations