Breast cancer is one of the most common and deadliest cancers among women. Since histopathological images contain sufficient phenotypic information, they play an indispensable role in the diagnosis and treatment of breast cancers. To improve the accuracy and objectivity of Breast Histopathological Image Analysis (BHIA), Artificial Neural Network (ANN) approaches are widely used in the segmentation and classification tasks of breast histopathological images. In this review, we present a comprehensive overview of the BHIA techniques based on ANNs. First of all, we categorize the BHIA systems into classical and deep neural networks for in-depth investigation. Then, the relevant studies based on BHIA systems are presented. After that, we analyze the existing models to discover the most suitable algorithms. Finally, publicly accessible datasets, along with their download links, are provided for the convenience of future researchers.

A Comprehensive Review for Breast Histopathology Image Analysis Using Classical and Deep Neural Networks

https://doi.org/10.1016/j.compbiomed.2022.105210

ECG-BiCoNet: An ECG-based pipeline for COVID-19 diagnosis using Bi-Layers of deep features integration

White blood cells (WBCs) are a portion of the immune system which fights against germs. Leukemia is the most common blood cancer which may lead to death. It occurs due to the production of a large number of immature WBCs in the bone marrow that destroy healthy cells. To overcome the severity of this disease, it is necessary to diagnose the shapes of immature cells at an early stage that ultimately reduces the modality rate of the patients. Recently different types of segmentation and classification methods are presented based upon deep-learning (DL) models but still have some limitations. This research aims to propose a modified DL approach for the accurate segmentation of leukocytes and their classification. The proposed technique includes two core steps: preprocessing-based classification and segmentation. In preprocessing, synthetic images are generated using a generative adversarial network (GAN) and normalized by color transformation. The optimal deep features are extracted from each blood smear image using pretrained deep models i.e., DarkNet-53 and ShuffleNet. More informative features are selected by principal component analysis (PCA) and fused serially for classification. The morphological operations based on color thresholding with the deep semantic method are utilized for leukemia segmentation of classified cells. The classification accuracy achieved with ALL-IDB and LISC dataset is 100% and 99.70% for the classification of leukocytes i.e., blast, no blast, basophils, neutrophils, eosinophils, lymphocytes, and monocytes, respectively. Whereas semantic segmentation achieved 99.10% and 98.60% for average and global accuracy, respectively. The proposed method achieved outstanding outcomes as compared to the latest existing research works.

/pdf/a-deep-network-designed-for-segmentation-and-classification-1ec7lvr2o5.pdf

A deep network designed for segmentation and classification of leukemia using fusion of the transfer learning models

Cancer is a severe condition of uncontrolled cell division that results in a tumor formation that spreads to other tissues of the body. Therefore, the development of new medication and treatment methods for this is in demand. Classification of microarray data plays a vital role in handling such situations. The relevant gene selection is an important step for the classification of microarray data. This work presents gene encoder, an unsupervised two-stage feature selection technique for the cancer samples’ classification. The first stage aggregates three filter methods, namely principal component analysis, correlation, and spectral-based feature selection techniques. Next, the genetic algorithm is used, which evaluates the chromosome utilizing the autoencoder-based clustering. The resultant feature subset is used for the classification task. Three classifiers, namely support vector machine, k-nearest neighbors, and random forest, are used in this work to avoid the dependency on any one classifier. Six benchmark gene expression datasets are used for the performance evaluation, and a comparison is made with four state-of-the-art related algorithms. Three sets of experiments are carried out to evaluate the proposed method. These experiments are for the evaluation of the selected features based on sample-based clustering, adjusting optimal parameters, and for selecting better performing classifier. The comparison is based on accuracy, recall, false positive rate, precision, F-measure, and entropy. The obtained results suggest better performance of the current proposal.

Gene encoder: a feature selection technique through unsupervised deep learning-based clustering for large gene expression data

Applying supervised contrastive learning for the detection of diabetic retinopathy and its severity levels from fundus images

This paper demonstrates the reversible logic synthesis for the n-to-2n decoder, where n is the number of data bits. The circuits are designed using only reversible fault tolerant Fredkin and Feynman double gates. Thus, the entire scheme inherently becomes fault tolerant. Algorithm for designing the generalized decoder has been presented. In addition, several lower bounds on the number of constant inputs, garbage outputs and quantum cost of the reversible fault tolerant decoder have been proposed. Transistor simulations of the proposed decoder are shown using standard p-MOS 901 and n-MOS 902 model with delay of 0.030 ns and 0.12 m channel length, which proved the functional correctness of the proposed circuits. The comparative results show that the proposed design is much better in terms of quantum cost, delay, hardware complexity and has significantly better scalability than the existing approach.

A Low Power Fault Tolerant Reversible Decoder Using MOS Transistors

Diabetic Retinopathy is a human eye disease that causes damage to the eye's retina and may ultimately result in complete blindness. Early detection of diabetic retinopathy is needed to avoid complete blindness. Physical tests, such as visual acuity test, dilation of pupils, optical consistency tomography, is used to detect diabetic retinopathy. However, it is costly in terms of time and might affect the patients. In these consequences, this paper detects the presence of Diabetic Retinopathy in the human eye using a machine learning algorithm. The proposed method applies classification algorithms on several features (e.g., optical disk diameter, lesion-specific (microaneurysms, exudates) or presence of hemorrhages) of an existing Diabetic Retinopathy dataset. Then the features were extracted and used for the final decision making to predict the presence of diabetic retinopathy. The proposed system used Decision Tree, Logistic Regression. Support Vector Machine for the prediction. The proposed method achieved 88% accurate results which is much better than the existing works. Moreover, the proposed method achieves a better score in precision and recall which are 97% and 92%, respectively compared to the existing result 72% and 63%, i.e., more the 25% in each category on average which proves the enormousness of the proposed method.

An Improved Approach for Detection of Diabetic Retinopathy Using Feature Importance and Machine Learning Algorithms

Leukemia originates in bone marrow. It massively affects the production of appropriate blood cells. Hence, its early detection is very crucial for human living. Generally, computational approaches for Leukemia detection use microscopic blood cells images. Then, machine learning based models are trained and tested for accurate measurement. The main challenge here is to achieve an acceptable accuracy with a scalable method. However, data inconsistency, missing values and data incompleteness made the researchers’ job much more difficult. In these consequences, this paper proposes a scalable Leukemia prediction method based on a publicly available ALL_IDB dataset using the HOG feature descriptor and Logistic Regression. Initially, the proposed method used Canny edge detector and noise reduction operators to detect the exact shape of Lymphocytes. Then, Principal Component Analysis (PCA) is applied to the detected image shapes. The PCA reduces the data dimensions without losing any valuable information and thus greatly minimizes the afterward computational cost. Finally, a classifier based model is produced for unforeseen events and it is tested. The results are validated using n-fold cross-validation technique, where n is a positive integer greater than or equal to three. The maximum average accuracy of the proposed model is 96% which is much higher than the state-of-the-art schemes.

http://dspace.ewubd.edu/bitstream/2525/2944/1/Hossain_Abedy.pdf

Leukemia Prediction from Microscopic Images of Human Blood Cell Using HOG Feature Descriptor and Logistic Regression

Breast cancer is one of the most common types of cancer among women. More than 1.5 million women each year are affected by breast cancer, which causes approximately 15% of all cancer deaths among women and the rate is increasing each year. Automatic detection of breast cancer could help to decrease the mortality rate by making diagnosis less time consuming and more accurate. Analysis of a breast cancer histopathology image is a challenging task. There have been many research studies on breast cancer tissue image classification. In this paper, we propose a machine learning model to automate the classification of benign and malignant tissue images. The best validation accuracy using this model for each of the magnification factor ranges from 92% to 94% and the best test accuracy ranges from 92% to 96%. The performance for each of the combination of feature extractor and classifier is also analyzed in terms of precision, recall, and f1-score, both in tabular form and graphically.

Md. Shamsujjoha

Papers

A Low Power Fault Tolerant Reversible Decoder Using MOS Transistors

An Improved Approach for Detection of Diabetic Retinopathy Using Feature Importance and Machine Learning Algorithms

Leukemia Prediction from Microscopic Images of Human Blood Cell Using HOG Feature Descriptor and Logistic Regression

Transfer Learning and Supervised Classifier Based Prediction Model for Breast Cancer

Design of a compact reversible fault tolerant field programmable gate array: A novel approach in reversible logic synthesis