An Ensemble Rule Learning Approach for Automated Morphological Classification of Erythrocytes

Abstract

The analysis of pathophysiological change to erythrocytes is important for early diagnosis of anaemia. The manual assessment of pathology slides is time-consuming and complicated regarding various types of cell identification. This paper proposes an ensemble rule-based decision-making approach for morphological classification of erythrocytes. Firstly, the digital microscopic blood smear images are pre-processed for removal of spurious regions followed by colour normalisation and thresholding. The erythrocytes are segmented from background image using the watershed algorithm. The shape features are then extracted from the segmented image to detect shape abnormality present in microscopic blood smear images. The decision about the abnormality is taken using proposed multiple rule-based expert systems. The deciding factor is majority ensemble voting for abnormally shaped erythrocytes. Here, shape-based features are considered for nine different types of abnormal erythrocytes including normal erythrocytes. Further, the adaptive boosting algorithm is used to generate multiple decision tree models where each model tree generates an individual rule set. The supervised classification method is followed to generate rules using a C4.5 decision tree. The proposed ensemble approach is precise in detecting eight types of abnormal erythrocytes with an overall accuracy of 97.81% and weighted sensitivity of 97.33%, weighted specificity of 99.7%, and weighted precision of 98%. This approach shows the robustness of proposed strategy for erythrocytes classification into abnormal and normal class. The article also clarifies its latent quality to be incorporated in point of care technology solution targeting a rapid clinical assistance.

Vision-based malaria parasite image analysis: a systematic review

“…This algorithm is helpful in predicting and identifying malaria-infected and non-infected parasite using relevant features. Maity et al (2017) gives a good prediction based on the disease symptoms than other earlier boosting algorithms (Maity et al, 2017). The ada-boost algorithm follows a procedure of setting the weights equally in the first round and then gradually, it increases the weights of the misclassified examples so that the weak learners are considered more over the hard learners.…”

“…The c4.5 classifier is mainly used as a statistical classifier to examine both dependent and discrete data. It has been used by Maity et al (2017) in image retrieval process for either single or multiple features (texture, shape, colour, etc. ).…”

“…Naive Bayes classifier is a supervised machine learning algorithm which comprises of many probabilistic classification techniques. This algorithm has been used by Maity et al (2017) for characterising malaria infection cells using multivariate Gaussian distribution. It has been observed that this classification technique helps in reducing the number of tests undertaken by the patients and is much more effective in detecting malaria disease.…”

“…It is a pattern recognition classification approach used to assign a task to each value in the sequence. Maity et al (2017) used this type of labeling concentrates in classifying each token and utilising information of neighbours tokens for labeling. LeCun et al (1998) proposed an ensemble learning algorithm which is been designed by taking together a set of classifiers and predictive models as Bagging, Boosting, Bayesian combination, many more techniques.…”

Comparison of Diagnostic Performance of Classification Trees and Hematological Discrimination Functions for Differential Diagnosis between the Two Common Hematologic Disorders: An Application of Multidimensional Scaling and Cluster Analysis

“…Recently, the accessibility of powerful statistical software programs has paved the way for the application of advanced statistical models such as data mining techniques in differential diagnosis. However, few studies have already used such advanced statistical methods and data mining techniques for differential diagnosis of hematological data (40,(57)(58)(59)(60)(61)(62)(63)(64)(65)(66). Therefore, this paper aims at comparing tree algorithms as powerful machine-learning methods with hematological indices in differentiation between IDA and βTT.…”