Acute Lymphoblastic Leukemia Cell Detection in Microscopic Digital Images Based on Shape and Texture Features

Tuba, Eva; Strumberger, Ivana; Bačanin, Nebojša; Živković, Dejan; Tuba, Milan

doi:10.1007/978-3-030-26354-6_14

Cited by 13 publications

(15 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The best results are printed in bold. Based on the results presented in Table 3, it can be concluded that the proposed GAO-based method achieved better accuracy compared to all other methods presented by Mishra et al (2017) and Tuba et al (2019). Sensitivity was also the best for the case where the proposed GAO-based classification method was used.…”

Section: Tp Tn Accuracy Tp Tn Fp Fnmentioning

confidence: 84%

“…The confusion matrix was used to compute the sensitivity, the specificity and the accuracy of the model. These results, achieved by the proposed algorithm, are presented in the last column in Table 3, along with the results presented by Mishra et al, (2017) and Tuba et al, (2019). The best results are printed in bold.…”

Section: Tp Tn Accuracy Tp Tn Fp Fnmentioning

confidence: 90%

“…For nucleus extraction, the thresholding technique was used. Based on the previous work presented in (Tuba et al, 2019) and repeated experiments, it was established that the best threshold value for extracting the nucleus is 0.35 when the saturation range is [0,1].…”

Section: Feature Extraction and Classificationmentioning

confidence: 99%

“…All the results reported by Mishra et al (2017) are also presented in this paper. Tuba et al (2019) used the same set of features, but the support vector machine was optimized by the bare bones fireworks algorithm (Li & Tan, 2018). The image dataset that was used is ALL-IDB2 dataset (Labati, Piuri & Scotti, 2011) that contains 130 blast and 130 normal lymphocyte images.…”

Section: Tp Tn Accuracy Tp Tn Fp Fnmentioning

confidence: 99%

“…To examine the contribution of the GAO algorithm alone, the simulation was repeated by the proposed GAO algorithm, but using the same set of feature as in (Tuba et al, 2019), where the BBFWA was used. The proposed GAO method, with the same features used in Tuba et al (2019), achieved a sensitivity 94.61%, a specificity 90.00% and a classification accuracy 92.31%, which improved the results presented in (Tuba et al, 2019): sensitivity for 0.40%, specificity for 0.63% and classification accuracy for 0.47%. This demonstrates a great potential of the novel GAO algorithm since in its first practical application, without any modifications, it outperformed well established state-of-the-art BBFWA.…”

Section: Tp Tn Accuracy Tp Tn Fp Fnmentioning

confidence: 99%

See 4 more Smart Citations

Generative Adversarial Optimization (GOA) for Acute Lymphocytic Leukemia Detection

Tuba¹,

Tuba²

2019

SIC

Self Cite

View full text Add to dashboard Cite

Computer-aided diagnostic systems represent an active research topic that captures the attention of scientists from various fields. Due to the complexity of human body and the various sources that can be used to generate medical digital images, numerous methods designed for analysing and detecting of different diseases have been developed. In this paper the focus is on detecting acute lymphocytic leukemia in microscopic digital blood images. A method for lymphocyte classification as normal cells or blasts is proposed. For purpose of classification, each cell is described by 11 features, namely 5 shape and 6 texture features. The support vector machine was used as a classifier and its parameters were tuned by means of a novel optimization algorithm: the generative adversarial optimization (GAO) algorithm. No other applications of this optimization algorithm-the latest so far-were found in literature. The proposed method for acute lymphocytic leukemia detection is compared with the Naïve Bayes classificatory, the k-nearest neighbour, the back propagation neural network, and the support vector machine optimized by the bare bones fireworks algorithm. The proposed GAO-based method achieved higher classification accuracy which exhibits the great potential of this new algorithm.

show abstract

Section: Tp Tn Accuracy Tp Tn Fp Fnmentioning

confidence: 84%

Section: Tp Tn Accuracy Tp Tn Fp Fnmentioning

confidence: 90%

Section: Feature Extraction and Classificationmentioning

confidence: 99%

Section: Tp Tn Accuracy Tp Tn Fp Fnmentioning

confidence: 99%

Section: Tp Tn Accuracy Tp Tn Fp Fnmentioning

confidence: 99%

See 3 more Smart Citations

Generative Adversarial Optimization (GOA) for Acute Lymphocytic Leukemia Detection

Tuba¹,

Tuba²

2019

SIC

Self Cite

View full text Add to dashboard Cite

show abstract

Intra‐nucleus mosaic pattern (InMop) and whole‐cell Haralick combined‐descriptor for identifying and characterizing acute leukemia blasts on single cell peripheral blood images

et al. 2023

View full text Add to dashboard Cite

Acute leukemia is usually diagnosed when a test of peripheral blood shows at least 20% of abnormal immature cells (blasts), a figure even lower in case of recurrent cytogenetic abnormalities. Blast identification is crucial for White Blood Cell (WBC) Counting, which depends on both identifying the cell type and characterizing the cellular morphology, processes susceptible of inter‐ and intraobserver variability. The present work introduces an image combined‐descriptor to detect blasts and determine their probable lineage. This strategy uses an Intra‐nucleus Mosaic pattern (InMop) descriptor that captures subtle nuclei differences within WBCs, and Haralick's statistics which quantify the local structure of both nucleus and cytoplasm. The InMop captures WBC inner‐nucleus structure by applying a multiscale Shearlet decomposition over a repetitive pattern (mosaic) of automatically‐segmented nuclei. As a complement, Haralick's statistics characterize the local structure of the whole cell from an intensity co‐occurrence matrix representation. Both InMoP and Haralick‐based descriptors are calculated using the b‐channel from Lab color‐space. The combined‐descriptor is assessed by differentiating blasts from non‐leukemic cells with support vector machine (SVM) classifiers and different transformation kernels, in two public and independent databases. The first database‐D1 (n=260) is composed of healthy and Acute Lymphoid Leukemia (ALL) single cell images, and second database‐D2 contains Acute Myeloid Leukemia (AML) blasts (n=3,294) and non‐blast (n=15,071) cell images. In a first experiment, blasts vs non‐blast differentiation is performed by training with a subset of D2 (n=6,588) and testing in D1 (n=260), obtaining a training AUC of 0.991±0.002 and AUC=0.782 for the independent validation. A second experiment automatically differentiates AML blasts (260 images from D2) from ALL blasts (260 images from D1), with an AUC of 0.93. In a third experiment, state‐of‐the‐art strategies, VGG16 and RESNEXT convolutional neural networks (CNN), separate blast from non‐blast cells in both databases. The VGG16 showed an AUC of 0.673 and the RESNEXT of 0.75. Reported metrics for all the experiments are area under the ROC curve (AUC), accuracy and F1‐score.This article is protected by copyright. All rights reserved.

show abstract

DeepLeukNet—A CNN based microscopy adaptation model for acute lymphoblastic leukemia classification

Saeed¹,

Kumar

Khuhro

et al. 2023

Multimed Tools Appl

View full text Add to dashboard Cite

Acute Lymphoblastic Leukemia Cell Detection in Microscopic Digital Images Based on Shape and Texture Features

Cited by 13 publications

References 30 publications

Generative Adversarial Optimization (GOA) for Acute Lymphocytic Leukemia Detection

Generative Adversarial Optimization (GOA) for Acute Lymphocytic Leukemia Detection

Intra‐nucleus mosaic pattern (InMop) and whole‐cell Haralick combined‐descriptor for identifying and characterizing acute leukemia blasts on single cell peripheral blood images

DeepLeukNet—A CNN based microscopy adaptation model for acute lymphoblastic leukemia classification

Contact Info

Product

Resources

About