J Korean Soc Med Inform Search


Healthc Inform Res > Volume 4(1); 1998 > Article
Journal of Korean Society of Medical Informatics 1998;4(1):113-121.
DOI: https://doi.org/10.4258/jksmi.1998.4.1.113    Published online June 30, 1998.
Study on the Development of Optimal Pattern Recognition System for Chromosome Karyotype Classification
K G Nam, S H Eom, Y H Chang, G R Jun, K S Lee
1Department of Electronic Engineering, Pusan National University, Korea.
2Department of Computer and Information Processing, Dong-Ju College, Korea.
3Department of Biomedical Engineering, Pusan National University Hospital, Korea.
4Department of Electrical Engineering, Dong-A University, Korea.

The framing and analysis of the chromosome karyogram, which requires specific cytogenetic knowledge, is most important in the cytogenetic part. To improve classification accuracy, we proposed an algorithm for the chromosome image reconstruction in the image preprocessing part and the optimal pattern classification method using the hierarchical multilayer artificial neural network (HMANN) to classify the chromosome karyotype. We reconstructed chromosome images of twenty normal human chromosome by the image reconstruction algorithm. The four morphological and ten density feature parameters were extracted from the 920 reconstructed chromosome images. The feature parameters were composed to four combined feature parameters. The HMANN pattern classifier to classify chromosome images were composed to the HMANN 1 and the HMANN 2. The HMANN 1 was classified a chromosome group, and the HMANN 2 was classified an each chromosome number. Each combined feature parameters of the ten human chromosome images were used to learn HMANN and the rest of them were used to classify the chromosome images. According to the above experimental results, therefore, the Feature 4 has been shown very excellent in the classification experiment among Feature 1 through Feature 4.

Key Words: Pattern Recognition, Reconstruction Method, Hierarchical Structure, Chromosome, Morphological Feature


Browse all articles >

Editorial Office
1618 Kyungheegung Achim Bldg 3, 34, Sajik-ro 8-gil, Jongno-gu, Seoul 03174, Korea
Tel: +82-2-733-7637, +82-2-734-7637    E-mail: hir@kosmi.org                

Copyright © 2024 by Korean Society of Medical Informatics.

Developed in M2community

Close layer
prev next