Classification of Skin Lesions By Tda Alongside Xception Neural Network

Elyasi, N.; Hosseini Moghadam, M.

doi:10.22044/jadm.2022.10948.2239

Document Type : Applied Article

Authors

N. Elyasi ¹
M. Hosseini Moghadam ²

¹ Department of math and computer Sciences, Kharazmi University, Tehran, Iran.

² Department of Engineering, Kharazmi University, Tehran, Iran.

https://doi.org/10.22044/jadm.2022.10948.2239

Abstract

In this paper, we use the topological data analysis (TDA) mapper algorithm alongside a deep convolutional neural network in order to classify some medical images.
Deep learning models and convolutional neural networks can capture the Euclidean relation of a data point with its neighbor data points like the pixels of an image and they are particularly good at modeling data structures that live in the Euclidean space and not effective at modeling data structures that live in the non-Euclidean spaces. Topological data analysis-based methods have the ability to not only extract the Euclidean, but also topological features of data.
For the first time in this paper, we apply a neural network as one of the filter steps of the Kepler mapper algorithm to classify skin cancer images. The major advantage of this method is that Kepler Mapper visualizes the classification result by a simplicial complex, where neural network increases the accuracy of classification. Furthermore, we apply TDA mapper and persistent homology algorithms to analyze the layers of Xception network in different training epochs. Also, we use persistent diagrams to visualize the results of the analysis of layers of the Xception network and then compare them by Wasserstein distances.

Keywords

20.1001.1.23225211.2022.10.3.4.6

References

[1] O. Reiter, V. Rotemberg, K. Kose, A. C. Halpern, “Artificial Intelligence in Skin Cancer,” Proc Natl Acad Sci U S A., 108, pp. 7265-70, 2011.

[2] M. E. Aktas, E. Akbas, A. E. Fatmaoui, “Persistence homology of networks: methods and applications,” Applied Network Science https://doi.org/10.1007/s41109-019-0179-3 4, (61), pp. 1-28, 2019.

[3] V. Buhrmester, D. Münch, M. Arens, “Analysis of Explainers of Black Box Deep Neural Networks for Computer Vision: A Survey,”, pp. 249-274, 2019, Available: arXiv:1911.12116.

[4] H. Edelsbrunner, “Persistent Homology in Image Processing,” Graph-Based Representations in Pattern Recognition. GbRPR 2013. Lecture Notes in Computer Science, vol 7877. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-38221-5_19

[5] M. Hosseini Moghadam, M. M. Pedram, "Topological Data Analysis for Classification of DeepSat-4 Dataset," 2020 10th International Symposium onTelecommunications (IST), 2020, pp. 246-250, doi:10.1109/IST50524.2020.9345829.

[6] N. Elyasi, M. Hosseini Moghadam, “An Introduction to a New Text Classification and Visualization for Natural Language Processing Using Topological Data Analysis,” 2019, available: arXiv preprint arXiv:1906.01726.

[7] S. Gholizadeh, , A. Seyeditabari, W. Zadrozny, “Topological Signature of 19th Century Novelists: Persistent Homology in Text Mining,” big data and cognitive computing, 2, pp. 1-33, 2018.

[8] H. Lee, H. Kang, M. K. Chung, B. Kim, D. Lee, “Persistent Brain Network Homology From the Perspective of Dendrogram,” IEEE TRANSACTIONS ON MEDICAL IMAGING., 31, pp. 2267-77, 2012.

[9] J. L. Nielson, J. Paquette, A. W. Liu, C. F. Guandique, C. A. Tovar, etal, “Topological data analysis for discovery in preclinical spinal cord injury and traumatic brain injury,” Nature Communications, 6, pp. 1-12, 2015.

[10] M. E. Sardiu, , J. M. Gilmore, B. Groppe, L. Florens, M. P.Washburn, “Identification of Topological Network Modules in Perturbed Protein Interaction Networks,” Scientific Reports, 7, pp. 1-13, 2017.

[11] P. Tschandl, C. Rosendahl, H. Kittler, “The HAM10000 dataset, a large collection of multi-source dermatoscopic images of common pigmented skin lesions,” Scientific data, 5(1), 1-9, 2018.

[12] https://medium.com/@zurister/depth-wise-convolution-and-depth-wise-separable-convolution-37346565d4ec.

[13] F. Chollet, “Xception: Deep learning with depthwise separable convolutions,” In Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 1251-1258, 2017.

[14] A. Zomorodian, G. Carlsson, “Computing Persistent Homology,” Discrete and Computational Geometry, 33, pp. 249-274, , 2005.

[15] P. Bubenik, P. A. Dłotko, “persistence landscapes toolbox for topological statistics,” J. Symbol. Comput., 78, 91–114, 2016.

[16] G. Singh, F. Mémoli, G. Carlsson, “Topological methods for the analysis of high dimensional data sets and 3d object recognition,” In Eurographics Symposium on Point-Based Graphics (eds Botsch, M., Pajarola, R.), 6, pp. 1-12, 2007.

[17]https://tmap.readthedocs.io/en/latest/how2work.html.

[18] `KeplerMapper', http://doi.org/10.5281/zenodo.1054444, accessed Jan 2019.

[19] A. Hekler, , J. S, Utikal, A. H. Enk, A. Hauschild, M. Weichenthal, R. C. Maron, ..., A. Thiem, “Superior skin cancer classification by the combination of human and artificial intelligence,” European Journal of Cancer, 120, 114, 2019.

[20] Y.M. Chung, C.S. Hu, A. Lawson, C. Smyth, “Topological approaches to skin disease image analysis,” In2018 IEEE International Conference on Big Data (Big Data), pp. 100-105, 2018 Dec 10.

[21] ISIC2018. Available online: https://challenge2018.isic-archive.com/ (accessed on 19 August 2018).

[22] M. Kurmanji and F. Ghaderi, “Hand Gesture Recognition from RGB-D Data using 2D and 3D Convolutional Neural Networks: a comparative study,” Journal of AI and Data Mining, 8, 2, pp. 177-188, 2020.

Classification of Skin Lesions By Tda Alongside Xception Neural Network

References

References

Volume 10, Issue 3July 2022Pages 333-344

Volume 10, Issue 3
July 2022
Pages 333-344