Chromosome Medial Axis Extraction Method Based on Graphic Geometry and Competitive Extreme Learning Machines Teams (CELMT) Classifier for Chromosome Classification
- 295 Downloads
Automated chromosome classification is a vital task in cytogenetics and has been a common pattern recognition problem. Numerous attempts were made in the past decade years to characterize chromosomes for the intention of medical and cancer cytogenetics research. This paper proposes a graphic geometry-based approach for medial axis extraction and Competitive Extreme Learning Machine Teams (CELMT) with further correction method for chromosome classification. The initial two medial axis points are determined firstly according to the length of the intercept line in different directions, and then the complete medial axis for feature extraction is drawn based on the initial two points. After that, a base classifier ELMi, j is trained to differentiate a pair class chromosome (i, j), a total number of 276 classifiers are trained. Each base classifier will give a label and the final label will be determined by majority voting and further correction rules. Based on the experiment results, the method proposed in this paper can precisely extract the medial axis and extract the features to recognize the chromosome, the classification accuracy by using CELMT can achieve an average value of 96.23% and the running time is much shorter than the other classification algorithms.
KeywordsKaryotype Cytogenetics Medial axis Extreme Learning Machine
This work was supported by the National Natural Science Foundation of China (61922072, 61876169, 61673404).
- 1.Nair, R.M., Remya, R., Sabeena, K.: Karyotyping techniques of chromosomes: a survey. Int. J. Comput. Trends Technol. 22(1) (2015) Google Scholar
- 3.Ventura, R., Khmelinskii, A., Sanches, J.M.: Classifier-assisted metric for chromosome pairing. In: 2010 Annual International Conference of the IEEE Engineering in Medicine and Biology, pp. 6729–6732. IEEE (2010)Google Scholar
- 6.Wu, Q., Castleman, K.R.: Automated chromosome classification using wavelet-based band pattern descriptors. In: Proceedings 13th IEEE Symposium on Computer-Based Medical Systems, CBMS 2000, pp. 189–194. IEEE (2000)Google Scholar
- 7.Minaee, S., Fotouhi, M., Khalaj, B.H.: A geometric approach to fully automatic chromosome segmentation. In: 2014 IEEE Signal Processing in Medicine and Biology Symposium (SPMB), pp. 1–6. IEEE (2014)Google Scholar
- 8.Moradi, M., Setarehdan, S., Ghaffari, S.: Automatic landmark detection on chromosomes’ images for feature extraction purposes. In: Proceedings of the 3rd International Symposium on Image and Signal Processing and Analysis, ISPA 2003, vol. 1, pp. 567–570. IEEE (2003)Google Scholar
- 19.Sharma, M., Vig, L., et al.: Automatic chromosome classification using deep attention based sequence learning of chromosome bands. In: 2018 International Joint Conference on Neural Networks (IJCNN), pp. 1–8. IEEE (2018)Google Scholar
- 20.Gagula-Palalic, S., Can, M.: Human chromosome classification using competitive neural network teams (CNNT) and nearest neighbor. In: IEEE-EMBS International Conference on Biomedical and Health Informatics (BHI), pp. 626–629. IEEE (2014)Google Scholar
- 21.Uttamatanin, R., Yuvapoositanon, P., Intarapanich, A., Kaewkamnerd, S., Tongsima, S.: Band classification based on chromosome shapes. In: The 6th 2013 Biomedical Engineering International Conference, pp. 1–5. IEEE (2013)Google Scholar
- 25.Faculty of computer science and mathematics. http://www.fim.uni-passau.de/en/faculty/former-professors/mathematical-stochastics/chromosome-image-data. Accessed 10 Oct 2019