Metric-based fuzzy rough sets for brain tumor magnetic resonance imaging classification

Abstract

Computer-aided diagnosis systems help physicians diagnose diseases accurately at an early stage by automating preprocessing, image enhancement, and feature extraction, thus increasing patient survival rates. In this paper, we introduce an algorithm that leverages metric-based fuzzy positive regions to address the degradation of feature quality in brain tumor magnetic resonance imaging caused by inappropriate image enhancement. Employing sliding window blocks, the algorithm performs overlapping segmentation of magnetic resonance images and evaluates the membership of these blocks to decision classes by metric-based fuzzy positive regions. Blocks with the highest fuzzy positive region values are selected for multiple enhancement rounds, forming a candidate set that is sequentially integrated back into the original image. Finally, the features of the locally enhanced images are analyzed using the fuzzy positive region to generate the optimal feature set. To validate the effectiveness of the proposed algorithm, the features extracted using this method are compared with those extracted directly from the original image, globally enhanced images, and locally enhanced images processed based on similar fuzzy positive regions. The experimental results demonstrate that the proposed algorithm significantly outperforms the other three methods in various evaluation metrics, including the confusion matrix, classification accuracy, and the kappa coefficient.

Keywords

• Image processing
• fuzzy rough sets
• metric
• fuzzy positive region
• magnetic resonance imaging

References

1. [1] A.I. Awad, Machine learning techniques for fingerprint identification: A short review, Lect. Notes Comput. Sci. 322, 524–531, 2012.
2. [2] A. Alsalihi, H.K. Aljobouri and E.A.K. Altameemi, GLCM and CNN deep learning model for improved MRI breast tumors detection, Int. J. Online Biomed. Eng. 18, 123–137, 2022.
3. [3] S. Banach, Théorie des opérations linéaires, Warsaw 1932.
4. [4] Z. Bonikowski, E. Bryniarski and U.Wybraniec-Skardowska, Extensions and intentions in the rough set theory, Inf. Sci. 107 (1-4), 149–167, 1998.
5. [5] S. Basak, R. Jia and C. Lei, Face recognition using fuzzy logic, Proc. IEEE Int. Conf. Inf. Autom. 1317–1322, 2018.
6. [6] S. Bhuvaji, A. Kadam, P. Bhumkar, S. Dedge and S. Kanchan, Brain tumor classification (MRI), Kaggle 2020.
7. [7] B. Bahuleyan, L. Raghuram, V. Rajshekhar and A.G. Chacko, To assess the ability of MRI to predict consistency of pituitary macroadenomas, Br. J. Neurosurg. 20 (5), 324–326, 2006.
8. [8] L. Chen and Q. Chen, A novel classification algorithm based on kernelized fuzzy rough sets, Int. J. Mach. Learn. Cybern. 11 (12), 2565–2572, 2020.

9. [9] G. Chen and Z.S. Chen, Regional classification of urban land use based on fuzzy rough set in remote sensing images, J. Intell. Fuzzy Syst. 38 (3), 3803–3812, 2020.
10. [10] V. Chirchi, E. Chirchi and K.E. Chirchi, Pattern matching for the iris biometric recognition system uses KNN and fuzzy logic classifier techniques, Int. J. Inf. Technol. 16 (4), 2937–2944, 2024.
11. [11] X.F. Cheng, X.J. Li and X.D. Ma, A method for battery fault diagnosis and early warning combining isolated forest algorithm and sliding window, Energy Sci. Eng. 11 (12), 4493–4504, 2023.
12. [12] Y. Chen, S.K. Park, Y. Ma and R.K. Ala, A new automatic parameter setting method of a simplified PCNN for image segmentation, IEEE Trans. Neural Netw. 22 (6), 880–892, 2011.
13. [13] J. Cheng, W. Yang, M. Huang, W. Huang, J. Jiang, Y. Zhou, R. Yang, J. Zhao, Y. Feng, Q. Feng and W. Chen, Retrieval of brain tumors by adaptive spatial pooling and fisher vector representation, PLoS ONE 11 (6), e0157112, 2016.
14. [14] Y. Du, Q. Hu, D.G. Chen and P. Ma, Kernelized fuzzy rough sets based yawn detection for driver fatigue monitoring, Fundam. Inform. 111 (1), 65–79, 2011.
15. [15] A. Eleyan and H. Demirel, Co-occurrence matrix and its statistical features as a new approach for face recognition, Turk. J. Electr. Eng. Comput. Sci. 19 (2), 285–295, 2011.
16. [16] E.A. El-Dahshan, H.M. Mohsen, K. Revett and A.M. Salem, Computer-aided diagnosis of human brain tumor through MRI: A survey and a new algorithm, Expert Syst. Appl. 41 (11), 5526–5545, 2014.
17. [17] S.F. Farahbakhshian and M.T. Ahvanooey, A new gene selection algorithm using fuzzy-rough set theory for tumor classification, Control Eng. Appl. Inform. 22 (1), 14–23, 2020.
18. [18] M. Fatima and M. Pasha, Survey of machine learning algorithms for disease diagnostic, J. Intell. Learn. Syst. Appl. 9 (1), 1–16, 2017.
19. [19] T. Feng, S. Zhang and J. Mi, The reduction and fusion of fuzzy covering systems based on the evidence theory, Int. J. Approx. Reason. 53 (1), 87–103, 2012.
20. [20] J. Goubault-Larrecq, Non-Hausdorff topology and domain theory, Cambridge University Press 2013.
21. [21] X. Gu, Research on pulse coupled neural networks and their applications, Science Press, Beijing, 2003.
22. [22] H.W. Goo and Y.S. Ra, Advanced MRI for pediatric brain tumors with emphasis on clinical benefits, Korean J. Radiol. 18 (1), 194–207, 2017.
23. [23] N. Haydar, K. Alyousef, U. Alanan, R. Issa, F. Baddour, Z. Al-shehabi and M.H. Al-janabi, Role of magnetic resonance imaging (MRI) in grading gliomas comparable with pathology: A cross-sectional study from Syria, Ann. Med. Surg. 82, 104679, 2022.
24. [24] R.M. Haralick, K. Shanmugam and I. Dinstein, Textural features for image classification, IEEE Trans. Syst. Man Cybern. SMC-3 (6), 610–621, 1973.
25. [25] G. Kuntimad and H.S. Ranganath, Perfect image segmentation using pulse coupled neural networks, IEEE Trans. Neural Netw. 10 (3), 591–598, 1999.
26. [26] L. Kong, J. Zhang, Y. Wang, C. Zhang and Q. Hu, Distributed feature selection for big data using fuzzy rough sets, IEEE Trans. Fuzzy Syst. 28 (5), 846–857, 2020.
27. [27] S. Liu, Study on medical image enhancement based on wavelet transform fusion algorithm, J. Med. Imaging Health Inform. 7 (2), 388–392, 2017.
28. [28] L.Y. Liu and X.J. Fan, The design of system to texture feature analysis based on gray level co-occurrence matrix, Appl. Mech. Mater. 727-728, 904–907, 2015.
29. [29] T. Lindblad and J.M. Kinser, Image processing using pulse-coupled neural networks, Springer, London, 1998.
30. [30] J. Lee, S.R. Pant and H. Lee, An adaptive histogram equalization based local technique for contrast preserving image enhancement, Int. J. Fuzzy Log. Intell. Syst. 15 (1), 35–44, 2015.
31. [31] L. Lei, F. Xi and S. Chen, Finger-vein image enhancement based on pulse coupled neural network, IEEE Access 7, 57226–57237, 2019.
32. [32] M.L. McHugh, Interrater reliability: The kappa statistic, Biochem. Med. 22 (3), 276–282, 2012.
33. [33] R. Membarth, O. Reiche, F. Hannig, J. Teich, M. Körner and W. Eckert, HIPAcc: A domain-specific language and compiler for image processing, IEEE Trans. Parallel Distrib. Syst. 27 (1), 210–224, 2016.
34. [34] R. Nie, M. He, J. Cao, D. Zhou and Z. Liang, Pulse coupled neural network based MRI image enhancement using classical visual receptive field for smarter mobile healthcare, J. Ambient Intell. Humaniz. Comput. 10 (8), 3345–3357, 2019.
35. [35] F.J. Provost, On applied research in machine learning, Proc. Conf. Appl. Res. Mach. Learn. 1–8, 1998.
36. [36] Y. Qu, Q. Fu, C. Shang, A. Deng, R. Zwiggelaar, M. George and Q. Shen, Fuzzy-rough assisted refinement of image processing procedure for mammographic risk assessment, Appl. Soft Comput. 91, 106230, 2020.
37. [37] Y. Qu, C. Shang, Q. Shen, N. MacParthaláin and W. Wu, Kernel-based fuzzy-rough nearest-neighbour classification for mammographic risk analysis, Int. J. Fuzzy Syst. 17 (3), 471–483, 2015.
38. [38] Y. Qu, G. Yue, C. Shang, L. Yang, R. Zwiggelaar and Q. Shen, Multi-criterion mammographic risk analysis supported with multi-label fuzzy-rough feature selection, Artif. Intell. Med. 100, 101722, 2019.
39. [39] U. Raghavendra, A. Gudigar, A. Paul, T.S. Goutham, M.A. Inamdar, A. Hegde, A. Devi, C.P. Ooi, R.C. Deo, P.D. Barua, F. Molinari, E.J. Ciaccio and U.R. Acharya, Brain tumor detection and screening using artificial intelligence techniques: Current trends and future perspectives, Comput. Biol. Med. 163, 107063, 2023.
40. [40] U. Reddy, B.V.R. Reddy and B.E. Reddy, Categorization & recognition of lung tumor using machine learning representations, Curr. Med. Imaging Rev. 15 (4), 405–413, 2019.
41. [41] C.K. Sirajuddeen, S. Kansal and R.K. Tripathi, Adaptive histogram equalization based on modified probability density function and expected value of image intensity, Signal Image Video Process. 14 (1), 9–17, 2020.
42. [42] J. Shi, Y. Lei, J. Wu and G. Jeon, Uncertain active contour model based on rough and fuzzy sets for auroral oval segmentation, Inf. Sci. 492, 72–103, 2019.
43. [43] M.M. Subashini and S.K. Sahoo, Pulse coupled neural networks and its applications, Expert Syst. Appl. 41 (8), 3965–3974, 2014.
44. [44] R. Sowiski and D. Vanderpooten, A generalized definition of rough approximations based on similarity, IEEE Trans. Knowl. Data Eng. 12 (3), 331–336, 2000.
45. [45] R.B. Vallabhaneni and V. Rajesh, Brain tumour detection using mean shift clustering and GLCM features with edge adaptive total variation denoising technique, Alex. Eng. J. 57 (4), 2387–2392, 2018.
46. [46] M.J. Warrens, Weighted kappa is higher than Cohen’s kappa for tridiagonal agreement tables, Stat. Methodol. 8 (2), 268–272, 2011.
47. [47] J. Watts, G.A. Box, A. Galvin, P.R. Brotchie, N.M. Trost and T.R. Sutherland, Magnetic resonance imaging of meningiomas: A pictorial review, Insights Imaging 5 (1), 113–122, 2014.
48. [48] Q. Wang and R.K. Ward, Fast image/video contrast enhancement based on weighted thresholded histogram equalization, IEEE Trans. Consum. Electron. 53 (2), 757–764, 2007.
49. [49] F.X. Wu and X.B. Zhang, An enhanced method of color image combined PCNN based on NSCT, J. Ambient Intell. Humaniz. Comput. 7 (6), 1597–1608, 2016.
50. [50] X. Xu, G. Wang, S. Ding, Y. Cheng and X. Wang, Pulse-coupled neural networks and parameter optimization methods, Neural Comput. Appl. 28 (1), 671–681, 2017.
51. [51] W. Yao, Y. She and L.X. Lu, Metric-based L-fuzzy rough sets: Approximation operators and definable sets, Knowl.-Based Syst. 163, 91–102, 2019.
52. [52] W. Yao, G.X. Zhang and C.J. Zhou, Real-valued hemimetric-based fuzzy rough sets and an application to contour extraction of digital surfaces, Fuzzy Sets Syst. 459, 201–219, 2022.
53. [53] W. Zhu, H. Jiang, E. Wang, Y. Hou, L. Xian and J. Debnath, X-ray image global enhancement algorithm in medical image classification, Discrete Contin. Dyn. Syst. Ser. S 12 (4-5), 1297–1309, 2019.
54. [54] X. Zhang, C. Mei, J. Li, Y. Yang and T. Qian, Instance and feature selection using fuzzy rough sets: A bi-selection approach for data reduction, IEEE Trans. Fuzzy Syst. 31 (6), 1981–1994, 2023.
55. [55] N. Zulpe and V.P. Pawar, GLCM textural features for brain tumor classification, Int. J. Comput. Sci. Inf. Technol. 3 (3), 4138–4141, 2012.
56. [56] D. Zhou, H. Zhou, C. Gao and Y. Guo, Simplified parameters model of PCNN and its application to image segmentation, Pattern Anal. Appl. 19 (4), 939–951, 2016.