Optimizing Capsule Network Performance with Enhanced Squash Function for Classification Large-Scale Bone Marrow Cells

Abstract

Capsule networks (CapsNet) have emerged as a promising architectural framework for various machine-learning tasks and offer advantages in capturing hierarchical relationships and spatial hierarchies within data. One of the most crucial components of CapsNet is the squash function, which plays a pivotal role in transforming capsule activations. Despite the success achieved by standard squash functions, some limitations remain. The difficulty learning complex patterns with small vectors and vanishing gradients are major limitations. Standard squash functions may struggle to handle large datasets. We improve our methodology to enhance squash functions to address these challenges and build on our previous research, which recommended enhancing squash functions for future improvements. Thus, high-dimensional, and complex data scenarios improve CapsNet’s performance. Enhancing CapsNet for complex tasks like bone marrow (BM) cell classification requires optimizing its fundamental operations. Additionally, the squash function affects feature representation and routing dynamics. Additionally, this enhancement improves feature representation, preserves spatial relationships, and reduces routing information loss. The proposed method increased BM data classification accuracy from 96.99% to 98.52%. This shows that our method improves CapsNet performance, especially in complex and large-scale tasks like BM cells. Comparing the improved CapsNet model to the standard CapsNet across datasets supports the results. The enhanced squash CapsNet outperforms the standard model on MNIST, CIFAR-10, and Fashion MNIST with an accuracy of 99.83%, 73%, and 94.66%, respectively. These findings show that the enhanced squash function improves CapsNet performance across diverse datasets, confirms its potential for real-world machine learning applications, and highlight the necessity for additional research.

Keywords

• Capsule networks
• Enhanced squash function
• Imbalanced dataset

References

1. Afriyie Y, Weyori BA, Opoku AA. 2022a. Classification of blood cells using optimized capsule networks. Neural Process Lett. 54: 4809–482.
2. Afriyie Y, Weyori BA, Opoku AA. 2022b. Comparative evaluation performances of capsule networks for complex image classification. J. Data Inf. Manag, 4(3–4): 267-276.
3. Agustin RI, Arif A, Sukorini U. 2021. Classification of immature white blood cells in acute lymphoblastic leukemia L1 using neural networks particle swarm optimization. Neural Comput Appl, 33(17): 10869-10880. doi: 10.1007/S00521-021-06245-7/TABLES/5.
4. Ananthakrishnan B, Shaik A, Akhouri S, Garg P, Gadag V, Kavitha MS. 2022. Automated bone marrow cell classification for haematological disease diagnosis using siamese neural network. Diagnostics (Basel), 13(1): 3390. doi: 10.3390/DIAGNOSTICS13010112.
5. Anupama MA, Sowmya V, Soman KP. 2019. Breast cancer classification using capsule network with preprocessed histology images. In: Proceedings of the IEEE International Conference on Communication and Signal Processing, ICCSP, April 4-6, Melmaruvathur, India, pp: 143-147.
6. Aydın Atasoy N, Al Rahhawi AFA. 2024. Examining the classification performance of pre-trained capsule networks on imbalanced bone marrow cell dataset. Int J Imaging Syst Technol, 34(3): e23067. doi: 10.1002/IMA.23067.
7. Balasubramanian K, Ananthamoorthy NP, Ramya K. 2022. An approach to classify white blood cells using convolutional neural network optimized by particle swarm optimization algorithm. Neural Comput Appl, 34(18): 16089-16101. doi: 10.1007/S00521-022-07279-1/TABLES/13.
8. Baghel N, Verma U, Nagwanshi KK. 2022. WBCs-Net: type identification of white blood cells using convolutional neural network. Multimed Tools Appl, 81(29): 42131-42147. doi: 10.1007/S11042-021-11449-Z/TABLES/11.

9. Bajer D, Zonc B, Dudjak M, Martinovic G. 2019. Performance analysis of SMOTE-based oversampling techniques when dealing with data imbalance. In: Proceedings of the International Conference on Systems, Signals, and Image Processing, June 5-7, Osijek, Croatia, pp:265-271.
10. Basnet J, Alsadoon A, Prasad PWC, Al Aloussi S, Alsadoon OH. 2020. A novel solution of using deep learning for white blood cells classification: enhanced loss function with regularization and weighted loss (ELFRWL). Neural Process Lett, 52(2): 1517-1553. doi: 10.1007/S11063-020-10321-9.
11. Baydilli YY, Atila Ü. 2020. Classification of white blood cells using capsule networks. Comput Med Imaging Graph, 80: 101699. doi:10.1016/J.COMPMEDIMAG.2020.101699.
12. Chawla NV, Bowyer KW, Hall LO, Kegelmeyer WP. 2011. SMOTE: synthetic minority over-sampling technique. J Artif Intell Res, 16: 321-357.
13. Chawla NV, Lazarevic A, Hall LO, Bowyer KW. 2003. SMOTEBoost: improving prediction of the minority class in boosting. In: Proceedings of the 7th European Conference on Principles and Practice of Knowledge Discovery in Databases, September 22-26, Cavtat-Dubrovnik, Croatia, pp: 107–119.
14. Chang S, Liu J. 2020. Multi-lane capsule network for classifying images with complex background. IEEE Access, 8: 79876-79886.
15. Confusion Matrix for Multi-Class Classification 2024. URL: https://www.analyticsvidhya.com/blog/2021/06/confusion-matrix-for-multi-class-classification/ (accessed date: June 13, 2024).
16. Çınar A, Tuncer SA. 2021. Classification of lymphocytes, monocytes, eosinophils, and neutrophils on white blood cells using hybrid Alexnet-GoogleNet-SVM. SN Appl Sci, 3(4): 1-11. doi: 10.1007/S42452-021-04485-9/TABLES/4.
17. Dayı B, Üzen H, Çiçek İB, Duman ŞB. 2023. A novel deep learning-based approach for segmentation of different type caries lesions on panoramic radiographs. Diagnostics, 13(2): 202. doi: 10.3390/DIAGNOSTICS13020202.
18. Dhal KG, Rai R, Das A, Ray S, Ghosal D, Kanjilal R. 2023. Chaotic fitness-dependent quasi-reflected Aquila optimizer for superpixel based white blood cell segmentation. Neural Comput Appl, 35(21): 15315-15332. doi: 10.1007/S00521-023-08486-0/TABLES/10.
19. do Rosario VM, Breternitz M, Borin E. 2021. Efficiency and scalability of multi-lane capsule networks (MLCN). J Parallel Distrib Comput, 155: 63-73. doi: 10.1016/J.JPDC.2021.04.010.
20. El Alaoui-Elfels O, Gadi T. 2021. EMG-CapsNet: Elu multiplication gate capsule network for complex images classification. In: Proceedings of the 13th International Conference on Soft Computing and Pattern Recognition (SoCPaR 2021), Online, December 15 – 17, pp: 97–108.
21. Elreedy D, Atiya AF. 2019. A comprehensive analysis of synthetic minority oversampling technique (SMOTE) for handling class imbalance. Inf Sci (N Y), 505: 32-64.
22. Fazeli S, Samiei A, Lee TD, Sarrafzadeh M. 2022. Beyond labels: visual representations for bone marrow cell morphology recognition, ArXiv, 111-117. doi: 10.1109/ichi57859.2023.00025.
23. Gautam A, Singh P, Raman B, Bhadauria H. 2017. Automatic classification of leukocytes using morphological features and Naïve Bayes classifier. In: Proceedings of the IEEE Region 10 Annual International Conference, TENCON, November 5-8, Penang, Malaysia, pp: 1023-1027. doi: 10.1109/TENCON.2016.7848161.
24. Girdhar A, Kapur H, Kumar V. 2022. Classification of white blood cells using convolution neural network. Biomed Signal Process Control, 71: 103156. doi: 10.1016/J.BSPC.2021.103156.
25. Ghosh M, Das D, Chakraborty C, Ray AK. 2010. Automated leukocyte recognition using fuzzy divergence. Micron, 41(7): 840-846. doi: 10.1016/J.MICRON.2010.04.017.
26. Ghosh A, Jana ND, Mallik S, Zhao Z. 2022. Designing optimal convolutional neural network architecture using differential evolution algorithm. Patterns, 3(9):100567. doi: 10.1016/j.patter.2022.100567.
27. Guo H, Viktor HL. 2004. Learning from imbalanced data sets with boosting and data generation: the DataBoost-IM approach. SIGKDD Explor, 6(1): 30-39.
28. Goswami D. 2019. Application of capsule networks for image classification on complex datasets.URL: https://hdl.handle.net/2142/105694 (accessed: May 26, 2024).
29. Ha Y, Du Z, Tian J. 2022. Fine-grained interactive attention learning for semi-supervised white blood cell classification. Biomed Signal Process Control, 75: 103611. doi: 10.1016/J.BSPC.2022.103611.
30. He H, Bai Y, Garcia EA, Li S. 2008. ADASYN: adaptive synthetic sampling approach for imbalanced learning. In: Proceedings of the International Joint Conference on Neural Networks, June 1-8, Hong Kong, China, pp:1322-1328.
31. Hegde RB, Prasad K, Hebbar H, Singh BMK. 2019a. Comparison of traditional image processing and deep learning approaches for classification of white blood cells in peripheral blood smear images. Biocybern Biomed Eng, 39(2): 382-392. doi: 10.1016/J.BBE.2019.01.005.
32. Hegde RB, Prasad K, Hebbar H, Singh BMK. 2019b. Feature extraction using traditional image processing and convolutional neural network methods to classify white blood cells: a study. Australas Phys Eng Sci Med, 42(2): 627-638. doi: 10.1007/S13246-019-00742-9/FIGURES/11.
33. Hosseini M, Bani-Hani D, Lam SS. 2022. Leukocytes image classification using optimized convolutional neural networks. Expert Syst Appl, 205: 117672. doi: 10.1016/J.ESWA.2022.117672.
34. Hoogi A, Wilcox B, Gupta Y, Rubin DL. 2019. Self-attention capsule networks for object classification. URL: https://arxiv.org/abs/1904.12483v2 (accessed: May 03, 2024).
35. Juanjuan W, Mantao X, Hui W, Jiwu Z. 2007. Classification of imbalanced data by using the SMOTE algorithm and locally linear embedding. In: Proceedings of the International Conference on Signal Processing, ICSP, November 16-20, Guilin, China, pp: 1741- 1745.
36. Keys RG. 1981. Cubic convolution interpolation for digital image processing. IEEE Trans Acoust, 29(6): 1153-1160.
37. Kingma DP, Ba JL. 2015. Adam: a method for stochastic optimization. In: Proceedings of 3rd International Conference on Learning Representations, ICLR, May 7-9, San Diego, pp:13.
38. Kutlu H, Avci E, Özyurt F. 2020. White blood cells detection and classification based on regional convolutional neural networks. Med Hypotheses, 135: 109472. doi: 10.1016/J.MEHY.2019.109472.
39. LeCun Y, Bengio Y, Hinton G. 2015. Deep learning. Nature, 521(7553): 436-444. doi: 10.1038/nature14539.
40. Liu Y, Fu Y, Chen P. 2019. WBCaps: a capsule architecture-based classification model designed for white blood cells identification. In: Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS, July 23-27, Berlin, Germany, pp:7027-7030. doi: 10.1109/EMBC.2019.8856700.
41. Long F, Peng JJ, Song W, Xia X, Sang J. 2021. BloodCaps: a capsule network-based model for the multiclassification of human peripheral blood cells. Comput Methods Programs Biomed, 202: 105972. doi: 10.1016/J.CMPB.2021.105972.
42. Iesmantas T, Alzbutas R. 2018. Convolutional capsule network for classification of breast cancer histology images. In: Proceedings of the 15th International Conference, ICIAR 2018, June 27–29, Póvoa de Varzim, Portugal, pp 869–876 DOI: 10.1007/978-3-319-93000-8_85
43. Maldonado S, López J, Vairetti C. 2019. An alternative SMOTE oversampling strategy for high-dimensional datasets. Appl Soft Comput, 76: 380-389.
44. Matek C, Krappe S, Münzenmayer C, Haferlach T, Marr C. 2021. Highly accurate differentiation of bone marrow cell morphologies using deep neural networks on a large image data set. Blood, 138(20): 1917-1927. DOI: 10.1182/blood.2020010568
45. Mensah PK, Weyori BA, Ayidzoe MA. 2021. Evaluating shallow capsule networks on complex images. Int J Inf Technol (Singapore), 13(3): 1047-1057. doi: 10.1007/S41870-021-00694-Y/FIGURES/6.
46. Mirmohammadi P, Ameri M, Shalbaf A. 2021. Recognition of acute lymphoblastic leukemia and lymphocytes cell subtypes in microscopic images using random forest classifier. Phys Eng Sci Med, 44(2): 433-441. doi: 10.1007/S13246-021-00993-5/TABLES/2.
47. Mohamed M, Far B, Guaily A. 2012. An efficient technique for white blood cells nuclei automatic segmentation. In: Proceeding of IEEE International Conference on Systems, Man, and Cybernetics (SMC), October 14-17, Seoul, Korea (South), pp: 220-225. doi: 10.1109/ICSMC.2012.6377703.
48. Muhammad A, Arserim M, Ömer T. 2023. Compare the classification performances of convolutional neural networks and capsule networks on the Coswara dataset. DUJE, 14(2): 265-271. doi: 10.24012/dumf.1270429.
49. Nair P, Doshi R, Keselj S. 2021. Pushing the limits of capsule networks. URL: http://arxiv.org/abs/2103.08074 (accessed: May 03, 2024).
50. Patil AM, Patil MD, Birajdar GK. 2021. White blood cells image classification using deep learning with canonical correlation analysis. IRBM, 42(5): 378-389. doi: 10.1016/J.IRBM.2020.08.005.
51. Patrick MK, Adekoya AF, Mighty AA, Edward BY. 2022. Capsule networks – a survey. J King Saud Univ Comput Inf Sci, 34(1): 1295-1310. doi: 10.1016/J.JKSUCI.2019.09.014.
52. Ren H, Su J, Lu H. 2019. Evaluating generalization ability of convolutional neural networks and capsule networks for image classification via Top-2 classification. URL: https://arxiv.org/abs/1901.10112v4 (accessed: May 10, 2024).
53. Reza MS, Ma J. 2019. Imbalanced histopathological breast cancer image classification with convolutional neural network. In: Proceedings of the 14th IEEE International Conference on Signal Processing (ICSP), August 12-16, 8, Beijing, China, pp: 619-624.
54. Rezatofighi SH, Soltanian-Zadeh H, Sharifian R, Zoroofi RA. 2009. A new approach to white blood cell nucleus segmentation based on gram-schmidt orthogonalization. In: Proceedings of the international Conference on Digital Image Processing, ICDIP, March 7 – 9, Bangkok, Thailand, pp:107-111. doi: 10.1109/ICDIP.2009.19.
55. Sabour S, Frosst N, Hinton GE. 2017. Dynamic routing between capsules. In: preceding of the 31st Conference on Neural Information Processing Systems (NIPS 2017), December 5 - 7, Long Beach, CA, USA, pp: 3859–3869. doi.org/10.48550/arXiv.1710.09829
56. Sengur A, Akbulut Y, Budak U, Comert Z. 2019. White blood cell classification based on shape and deep features. In: Proceeding of International Conference on Artificial Intelligence and Data Processing Symposium, IDAP, September 21-22, Malatya, Turkey, pp 1-4.
57. Singh R, Ahmed T, Kumar A, Singh AK, Pandey AK, Singh SK. 2021. Imbalanced breast cancer classification using transfer learning. IEEE/ACM Trans Comput Biol Bioinform, 18(1): 83-93. doi: 10.1109/TCBB.2020.2980831.
58. Shiri P, Baniasadi A. 2021. Convolutional fully connected capsule network (CFC-CapsNet). In : Proceeding of the International Conference of Workshop on Design and Architectures for Signal and Image Processing (DASIP) - 14th edition, January 18-20, Budapest, Hungary, pp:19.
59. Shiri P, Sharifi R, Baniasadi A. 2020. Quick-CapsNet (QCN): a fast alternative to capsule networks. In: Proceedings of IEEE/ACS International Conference on Computer Systems and Applications, AICCSA, November 2-5, Antalya, Turkey, pp:1-7. doi: 10.1109/AICCSA50499.2020.9316525.
60. Somuncu E, Atasoy Aydın N. 2021. Realization of character recognition application on text images by convolutional neural network. J FAC ENG ARCHIT GAZ, 37(1):17-27.doi: 10.17341/gazimmfd.866552.
61. Stock W, Hoffman R. 2000. White blood cells 1: non-malignant disorders. Lancet, 355(9212): 1351-1357. doi: 10.1016/S0140-6736(00)02125-5.
62. Tamang T, Baral S, Paing MP. 2022. Classification of white blood cells: a comprehensive study using transfer learning based on convolutional neural networks. Diagnostics, 12(12).
63. Tasdelen A, Ugur AR. 2021. Artificial intelligence research on COVID-19 pandemic: a bibliometric analysis. In: Proceedings of 5th International Symposium on Multidisciplinary Studies and Innovative Technologies, ISMSIT, October 21-23, pp: 693-699. doi: 10.1109/ISMSIT52890.2021.9604573.
64. Vigueras-Guillén JP, Patra A, Engkvist O, Seeliger F. 2021. Parallel capsule networks for classification of white blood cells. In: preceding of the International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), September 27 to October 1, Strasbourg, France, pp: 743-752. doi.org/10.48550/arXiv.2108.02644
65. Xi E, Bing S, Jin Y. 2017. Capsule network performance on complex data. URL: https://arxiv.org/abs/1712.03480v1 (accessed: May 16, 2024).
66. Xiang C, Zhang L, Tang Y, Zou W, Xu C. 2018. MS-CapsNet: a novel multi-scale capsule network. IEEE Signal Process Lett, 25(12): 1850-1854.
67. Yao X, Sun K, Bu X, Zhao C, Jin Y. 2021. Classification of white blood cells using weighted optimized deformable convolutional neural networks. Artif Cells Nanomed Biotechnol, 49(1): 147-155. doi: 10.1080/21691401.2021.1879823.
68. Zhao L, Huang L. 2019. Exploring dynamic routing as a pooling layer. In: Proceeding of IEEE/CVF International Conference on Computer Vision Workshop (ICCVW),October 27-28, Seoul, Korea, pp: 738 – 742.
69. Zhao Z, Kleinhans A, Sandhu G, Patel I, Unnikrishnan KP. 2019. Capsule networks with max-min normalization. URL:http://arxiv.org/abs/1903.09662 (accessed: May 26, 2024).