Automated Approach for the Identification and Classification of Red Blood Cells Using LDA

Abstract

this study presents an automated approach for the identification and classification of red blood cells (RBC) in microscopic blood smear images. The primary goal of this research is to develop an efficient and accurate method for identifying and classifying red blood cells using a combination of Discrete Wavelet Transform (DWT) feature extraction technique and Linear Discriminant Algorithm (LDA) classifier to improve the accuracy and efficiency of traditional image processing techniques used for diagnosing various hematological disorders. The methodology involves preprocessing microscopic blood smear images, segmenting RBCs, and applying discrete wavelet transform feature extraction technique to acquire important morphological properties such as diameter, shape geometric factor, central pallor, and target flag for classification purposes. The extracted DWT features are then fed as input to LDA which employs a linear decision boundary to classify RBCs into normal (Normocytic) and abnormal (Microcytic) categories. The proposed method achieved a notable 85% accuracy with the LDA classifier.

Introduction

Blood analysis, especially examining red blood cells (RBCs), is vital for diagnosing various blood-related disorders. RBCs vary in size, shape, and concentration, with normocytic (normal size) and microcytic (smaller size) RBCs being key classifications used in diagnosing conditions like anemia. Traditional manual blood cell analysis is slow and imprecise, prompting the development of automated classification methods.

This study proposes an automated system for classifying microscopic blood smear images of RBCs using image processing and pattern recognition techniques. The system captures high-resolution images of blood smears, preprocesses them by converting to binary images, segments cells, and extracts features such as cell size and shape. Wavelet transform (2D-DWT) is applied to extract detailed features distinguishing normal and abnormal cells.

The classification is done using a Linear Discriminant Analysis (LDA) classifier, which divides cells into normocytic or microcytic categories based on extracted features. The system was tested on a dataset of 40 samples, achieving an accuracy of 85%, indicating promising results for improving diagnostic speed and accuracy while reducing manual effort.

Conclusion

This proposed system successfully demonstrates an automated approach for identifying and classifying red blood cells in microscopic blood smear images using Discrete Wavelet Transform (DWT) for feature extraction and the Linear Discriminant Algorithm (LDA) for classification. By leveraging important morphological properties such as diameter, shape geometric factor, central pallor, and target flag, the proposed method achieves an accuracy of 85%. The results indicate that the integration of DWT and LDA enhances classification efficiency compared to traditional image processing techniques.

References

[1] S. Diez-Silva, M., Dao, M., Han, J., Lim, C.-T., & Suresh, S. (2010). Shape and Biomechanical Characteristics of Human Red Blood Cells in Health and Disease. In MRS Bulletin (Vol. 35, Issue 5, p. 382). Springer Nature. https://doi.org/10.1557/mrs2010.571 [2] Huda Abdul-Hussain Obeed, Zainab Haider Ameen, Mohammed A. Fadhel a,c. , Laith Alzubaidi, “Automated Real-Time Sickle Cell Anemia Diagnosis Based On An Field Programmable Gate Array Accelerator”, Journal Of Southwest Jiaotong University, Vol. 54 No. 6 Dec. 2019, ISSN: 0258-2724 [3] Satish R.Suryawanshi, S. S. Sonone, “Disease Detection By Abnormalities In Erythrocyte Using Image Processing”, International Journal Of Scientific Research, Volume - 11 | Issue - 10 | October - 2022 | PRINT ISSN No. 2277 - 8179 | DOI : 10.36106/ijsr [4] Vaibhav V. Kamble, Rajendra D. Kokate, “Review on Multiple Classifier System in Pattern Recognition”, International Journal of Innovative Research in Science, Engineering and Technology, Volume 6, Special Issue 1, January 2017, ISSN (Online) : 2319 – 8753. [5] Supriya S. Pandarge, Varsha R.Ratnaparkhe, “Image Processing based Linear Discriminant and Quadratic Discriminant Classifier for Feature Extraction Models”, Proceedings of the International Conference on Intelligent Computing and Control Systems (ICICCS 2019) IEEE Xplore Part Number: CFP19K34-ART; ISBN: 978-1-5386-8113-8 [6] Manuel Gonz´alez-Hidalgo, F. A. Guerrero-Pe˜na, S. Herold-Garc´?a, Antoni Jaume-i-Cap´o, and P. D. Marrero-Fern´andez, “Red Blood Cell Cluster Separation from Digital Images for use in Sickle Cell Disease”, DOI 10.1109/JBHI.2014.2356402, IEEE Journal of Biomedical and Health Informatics. [7] A. Kavitha, S. Ramakrishnan, “Assessment of Human Red Blood Cell Aggregation Using Image Processing and Wavelets”, MEASUREMENT SCIENCE REVIEW, Volume 7, Section 2, No. 5, 2007. [8] Kamarul Hawari Ghazali, Mohd Fais Mansor, Mohd. Marzuki Mustafa and Aini Hussain, “Feature Extraction Technique using Discrete Wavelet Transform for Image Classification”, The 5th Student Conference on Research and Development -SCOReD 2007,11-12 December 2007, Malaysia, 1-4244-1470-9/07/$25.OO ©D2007 IEEE.

Copyright

Copyright © 2025 Satish R. Suryawanshi, Shaikh Aamer. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.