Table 2.

Comparison of different acute lymphoblastic leukaemia detection methods.

Authors, year	Method	Dataset	Preprocessing	Segmentation	Features extraction	Classification	Performance%
Mishra et al., 2017	Gray level cooccurrence matrix and random forest based acute lymphoblastic leukaemia detection [13]	Public ALL-IDB 1 (108 images)	Histogram equalization, Weiner filtering	Sobel, Perwitt, Marker based watershed segmentation	GLCM (gray level cooccurrence matrix), PPCA (Probabilistic Principal Component Analysis)	RF (random forest)	Accuracy 96.29%
Karthikeyan and Poornima, 2017	Microscopic image segmentation using fuzzy C-means for leukaemia diagnosis [14]	Google (19 images)	Histogram equalization, Median filter	K-means clustering, fuzzy c-means clustering algorithm	Gabor texture extraction method	SVM (support vector machine)	Accuracy 90%
Rawat et al., 2017	Classification of acute lymphoblastic leukaemia using hybrid hierarchical classifiers [15]	Public ALL-IDB 2 (260 images)	Histogram equalization, Order statistic filter	Global thresholding, morphological opening	Geometrical features, chromatic features, statistical texture features	Hybrid hierarchical classifiers kNN, PNN, SVM, SSVM, and ANFIS	Accuracy 99.2%
Joshi et al., 2013	White blood cells segmentation and classification to detect acute leukaemia [16]	Public ALL-IDB 1 (108 images)	Contrast Stretching and Histogram equalization	Otsu's threshold method	Shape features (area, perimeter, and circularity)	KNN (K-nearest neighbor)	Accuracy 93%
Putzu and Ruberto, 2013	White blood cells identification and classification from leukaemia blood image [17]	Public ALL-IDB 1 (108 images)	Histogram equalization	Triangle threshold method using Zack algorithm	Shape based features (area, perimeter, etc.), GLCM features	SVM (support vector machine)	Accuracy 92%
Li et al., 2016	Segmentation of white blood cell from acute lymphoblastic leukaemia images using dual-threshold method [18]	Public ALL-IDB (130 images)	Global Contrast Stretching	Dual-threshold segmentation	Binarization, morphological erosion, median filtering (postprocessing)	Not mentioned	Accuracy 98%
Amin et al., 2015	Recognition of acute lymphoblastic leukaemia cells in microscopic images using K-means clustering and support vector machine classifier [19]	Isfahan Al-Zahra and Omid Hospital pathology laboratories (146 ALL images and 166 lymphocytes images)	Histogram equalization	K-means clustering	Shape based features (area, perimeter, solidity, and eccentricity), histogram-based features (mean, standard deviation skewness, entropy, etc.)	SVM (support vector machine)	Accuracy 95.6%
Savita Dumyan, 2017	An enhanced technique for lymphoblastic cancer detection using artificial neural network [20]	Blood sample images (36 images)	Histogram equalization	Image binarization, canny edge detection technique	Shape based features, texture features, statistical features, moment invariants	Artificial neural network (ANN)	Accuracy 97.8%
Chatap and Shibu, 2014	Analysis of blood samples for counting leukaemia cells using support vector machine and nearest neighbour [21]	Public ALL-IDB 1 (108 images) and ALL-IDB 2 (260 images)	Histogram Equalization, Contrast Stretching	Otsu's threshold method	Shape based features (area, perimeter, and circularity)	K-nearest neighbor (KNN)	Accuracy 93%
Amin et al., 2015	Recognition of acute lymphoblastic leukaemia cells in microscopic images using K-means clustering and support vector machine classifier [22]	Isfahan Al-Zahra and Omid Hospital pathology laboratories (312 images)	Histogram Equalization, Linear Contrast Stretching	K-means clustering	Geometric or shape based (area, perimeter, convex, and solidity), first- and second-order statistical features	SVM (support vector machine)	Accuracy 97% (blast and normal), 95.6% (subtypes classification)
Patel and Mishra, 2015	Automated leukaemia detection using microscopic images [23]	Not mentioned	Median Filtering, Wiener Filtering	K-means clustering	Color features, geometric, texture, and statistical features	SVM (support vector machine)	Accuracy 93.57%
MoradiAmin et al., 2016	Computer aided detection and classification of acute lymphoblastic leukaemia cell subtypes based on microscopic image analysis [24]	Isfahan Al-Zahra and Omid Hospital pathology laboratories (312 images)	Histogram equalization	Fuzzy C-means, watershed algorithm	Geometric or shape based (area, perimeter, convex, and solidity), first- and second-order statistical features	SVM (support vector machine)	Accuracy 97.52%
Mohapatra and Patra, 2010	Automated cell nucleus segmentation and acute leukaemia detection in blood microscopic images [25]	University of Virginia, Ispat General Hospital, Rourkela, Odisha (108 images)	Selective median filtering, Unsharp Masking	K-means clustering, nearest neighbor	Fractal dimension, shape features including contour signature and texture, color features	SVM (support vector machine)	Accuracy 95%
Mohapatra et al., 2011	Fuzzy based blood image segmentation for automated leukaemia detection [26]	University of Virginia, Ispat General Hospital, Rourkela, Odisha (108 images)	Selective median filtering, Unsharp Masking	Gustafson Kessel clustering, nearest neighbor	Fractal dimension, shape features including contour signature and texture, color features	SVM (support vector machine)	Accuracy 93%
Mohapatra et al., 2014	An ensemble classifier system for early diagnosis of acute lymphoblastic leukaemia in blood microscopic images [27]	Ispat General Hospital, Rourkela, Odisha (150 images)	Contrast enhancement, Selective median filtering	Shadowed C-means (SCM) clustering	Fractal dimension, shape features including contour signature and texture, color features	Ensemble method (Naive Bayesian, K-nearest neighbor, multilayer perceptron, radial basis functional neural network, support vector machines)	Accuracy 94.73%
Samadzadehaghdam et al., 2015	Enhanced recognition of acute lymphoblastic leukaemia cells in microscopic images based on feature reduction using principle component analysis [28]	Isfahan Al-Zahra and Omid Hospital pathology laboratories (21 Images)	Histogram equalization	Fuzzy C-means, watershed algorithm	Geometric or shape based (area, perimeter, convex, and solidity), statistical features	SVM (support vector machine)	Accuracy 96.33%
Putzua et al., 2017	Leucocyte classification for leukaemia detection using image processing techniques [29]	Public ALL-IDB 1 (108 images) and ALL-IDB 2 (260 images)	Histogram equalization and contrast stretching	Zack algorithm	Shape features, color features, texture features	SVM (support vector machine)	Accuracy 92%
Sadeghian et al., 2009	A framework for white blood cell segmentation in microscopic blood images using digital image processing [30]	L2 type ALL blood images (20 images)	Gaussian filter, Standard deviation	Canny edge detection technique, Zack algorithm	Not mentioned	Not mentioned	Accuracy 92% (nucleus segmentation), 78% (cytoplasm segmentation)
Mohapatra and Patra, 2010	Automated leukaemia detection using Hausdorff dimension in blood microscopic images [31]	University of Virginia, Ispat General Hospital, Rourkela, Odisha (108 images)	Selective median filtering, Unsharp masking	K-means clustering	Hausdorff dimension, shape features, color features	SVM (support vector machine)	Accuracy 95%
Mohapatra et al., 2010	Image analysis of blood microscopic images for acute leukaemia detection [32]	University of Virginia, Ispat General Hospital, Rourkela, Odisha (108 images)	Selective median filtering, Unsharp masking	Fuzzy C-means clustering, nearest neighbor	Fractal features (Hausdorff dimension), shape features, contour signature, color, texture features	SVM (support vector machine)	Accuracy 95%