Image segmentation is the most important step and a key technology in image processing which directly affect the next processing. In human blood cell segmentation cases, many methods were applied for obtaining better results. It is basically an improved visualization to observe blood cell under blood smear process. This paper will present an approach for red blood cell (RBC) segmentation which is a part of study to perform automated counting for RBC. The methods involve are Ycbcr color conversion, masking, morphological operators and watershed algorithm. The combination of Ycbcr color conversion and morphological operator produce segmented white blood cell (WBC) nucleus. Then it is being used as a mask to remove WBC from the blood cell image. Morphological operators involve binary erosion diminish small object like platelet. The resulted RBC segmentation is passing through marker controlled watershed algorithm which handles overlapping cells. The improvement need to be done for both segmentation and overlapped cell handling to obtain better result in the future.

Red blood cell segmentation using masking and watershed algorithm: A preliminary study

The k-nearest neighbor (k-NN) is a traditional method and one of the simplest methods for classification problems. Even so, results obtained through k-NN had been promising in many different fields. Therefore, this paper presents the study on blasts classifying in acute leukemia into two major forms which are acute myelogenous leukemia (AML) and acute lymphocytic leukemia (ALL) by using k-NN. 12 main features that represent size, color-based and shape were extracted from acute leukemia blood images. The k values and distance metric of k-NN were tested in order to find suitable parameters to be applied in the method of classifying the blasts. Results show that by having k = 4 and applying cosine distance metric, the accuracy obtained could reach up to 80%. Thus, k-NN is applicable in the classification problem.

Classification of blasts in acute leukemia blood samples using k-nearest neighbour

Colour image segmentation has becoming more popular for computer vision due to its important process in most medical analysis tasks. One of the main tasks is the segmentation of white blood cell (WBC) where the WBC composition reveals important diagnostic information of a patient. In this paper, the combination between linear contrast technique and colour segmentation based on HSI (Hue, Saturation, Intensity) colour space were used in order to obtain a fully segmented abnormal WBC and nucleus of acute leukaemia images. The unsupervised segmentation technique namely k-means clustering algorithm is used to ease the segmentation process. By implementing the proposed segmentation technique, the fully segmented WBC which consists of cytoplasm and nucleus regions can be achieved by using the combination of linear contrast technique and segmentation based on H component image. Meanwhile, the fully segmented nucleus can be obtained by applying the segmentation based on S component image. The combinations between linear contrast technique and segmentation based on HSI colour space have produced a better effect on improving the accuracy of WBC segmentation with segmentation accuracies of 99.02% and 99.05% for segmented WBC and nucleus, respectively.

Unsupervised colour segmentation of white blood cell for acute leukaemia images

Contrast enhancement and image segmentation play an important process in most medical image analysis tasks. One of the main tasks is the analyzing of white blood cells (WBC) where the WBC composition reveals important diagnostic information of a patient. This paper presents a two phase methodology in order to obtain a fully segmented abnormal white blood cell (blast) and nucleus in acute leukaemia images. In the first phase, the three contrast enhancement techniques which are partial contrast, bright stretching and dark stretching were used to improve the image quality. Contrast enhancement techniques enhanced the area of interest of acute leukaemia for easing the segmentation process. In the second phase, image segmentation based on HSI (Hue, Saturation, Intensity) colour space is proposed. The proposed technique helps to improve the image visibility and has successfully segmented the acute leukaemia images into two main components: blast and nucleus. The combination between contrast enhancements and image segmentation has good effect on improving the accuracy of segmentation. Hence, information gain from the resultant images would become useful for haematologists to further analysis the types of acute leukaemia.

Improving colour image segmentation on acute myelogenous leukaemia images using contrast enhancement techniques

Leukaemia is a cancer of blood that causes more death than any other cancers among children and young adults under the age of 20. This disease can be cured if it is detected and treated at the early stage. Based on this argument, the requirement for fast analysis of blood cells for leukaemia is of paramount importance in the healthcare industry. This paper presents the classification of white blood cells (WBC) inside the Acute Lymphoblastic Leukaemia (ALL) and Acute Myelogenous Leukaemia blood samples by using the Multilayer Perceptron (MLP) and Simplified Fuzzy ARTMAP (SFAM) neural networks. Here, the WBC will be classified as lymphoblast, myeloblast and normal cell for the purpose of categorization of acute leukaemia types. Two different training algorithms namely Levenberg-Marquardt and Bayesian Regulation algorithms have been employed to train the MLP network. There are a total of 42 input features that consist of the size, shape and colour based features, have been extracted from the segmented WBCs, and used as the neural network inputs for the classification process. The classification results indicating that all networks have produced good classification performance for the overall proposed features. However, the MLP network trained by Bayesian Regulation algorithm has produced the best classification performance with testing accuracy of 95.70% for the overall proposed features. Thus, the results significantly demonstrate the suitability of the proposed features and classification using MLP and SFAM networks for classifying the acute leukaemia cells in blood sample.

Classification of Acute Leukaemia Cells using Multilayer Perceptron and Simplified Fuzzy ARTMAP Neural Networks

Fast and cost-effective production of blood cell count reports is of paramount importance in the healthcare industry. The traditional method of manual count under the microscope yields inaccurate results and put an intolerable amount of stress on the Medical Laboratory Technicians. Although there are hardware solutions such as the Automated Hematology Counter, developing countries like Sri Lanka are not capable of deploying such prohibitively expensive machines in every hospital laboratory in the country. As a solution to this problem, this research project aims to provide a software-based cost effective and an efficient alternative in recognizing and analyzing blood cells.

Gpm Priyanka

Papers

An extensible computer vision application for blood cell recognition and analysis

Ensemble learning-based deep neural network model for face recognition

Deep learning based video surveillance for predicting vehicle density in real time scenario