Breast Cancer Detection in Saudi Arabian Women Using Hybrid Machine Learning on Mammographic Images

Breast cancer (BC) is the most common cause of women’s deaths worldwide. The mammography technique is the most important modality for the detection of BC. To detect abnormalities in mammographic images, the Breast Imaging Reporting and Data System (BI-RADs) is used as a baseline. The correct allocation of BI-RADs categories for mammographic images is always an interesting task, even for specialists. In this work, to detect and classify the mammogram images in BI-RADs, a novel hybrid model is presented using a convolutional neural network (CNN) with the integration of a support vector machine (SVM). The dataset used in this research was collected from different hospitals in the Qassim health cluster of Saudi Arabia. The collection of all categories of BI-RADs is one of the major contributions of this paper. Another significant contribution is the development of a hybrid approach through the integration of CNN and SVM. The proposed hybrid approach uses three CNN models to obtain ensemble CNN model results. This ensemble model saves the values to integrate them with SVM. The proposed system achieved a classification accuracy, sensitivity, specificity, precision, and F1-score of 93.6%, 94.8%, 96.9%, 96.6%, and 95.7%, respectively. The proposed model achieved better performance compared to previously available methods.     

基于支持向量机的印品缺陷分类方法

目的研究印品图像的各类形状缺陷,建立基于支持向量机(Support vector machine,SVM)的印品形状缺陷分类模型。方法对印品进行符合人眼视觉特性的缺陷识别,并对提取缺陷进行特征分析。将特征数据导入支持向量机进行训练学习,SVM分类器对缺陷图像进行测试。结果分类器对点缺陷和面缺陷的识别率为100%,对线缺陷的分类准确率达93.94%。结论基于SVM的缺陷分类方法能较好地满足印品质量检测的需求。     

Correspondence analysis and establishment of evaluation model of classification performance indices for a turbo air classifier

The operational parameters of a turbo air classifier including feeding speed, rotor cage's rotary speed and air inlet velocity affect its classification performance directly, such as cut size, classification precision, classification efficiency, fine powder yield, particle fineness and degree of dispersion. Current methods of optimizing operational parameters and improving the classification performance of a turbo air classifier are almost single objective decision only for one of the classification performance indices. In this paper, the multi‐objective programming (MOP) model on classification performance for a turbo air classifier is established to evaluate these performance indices comprehensively and achieve optimal classification performance. To minimize the effect of repeatability within these classification performance indices, correspondence analysis is applied to determine the evaluation indices of this MOP model. According to correspondence analysis on the fine talc classification experimental data as well as the calcium carbonate classification experimental data, there is a very strong correlation between cut size and D90; there is also a very strong correlation between cut size and fine powder yield. Thus D90 and fine powder yield are filtered out and they aren't discussed in the evaluation model. The variation coefficient method is introduced to calculate weights of the evaluation function, and the dimensionless transformation method is used to eliminate the effects of different dimension. Thus, the optimal solution among the experimental data is obtained through solving the evaluation function. For the talc classification experiments, the optimal operational parameter combinations are: the feeding speed is 40 kg · h^–1, the air inlet velocity is 5 m · s^–1 and the rotor cage's rotary speed is 1200 ? min^–1. The classification performance indices are: cut size is 16.5 μm, classification precision index is 0.59, Newton classification efficiency is 57% and degree of dispersion is 2.13. For the calcium carbonate classification experiments, the optimal operational parameter combinations are: the feeding speed is 92 kg · h^–1, the air inlet velocity is 12 m · s^–1 and the rotor cage's rotary speed is 1200 ? min^–1. The classification performance indices are: cut size is 31.4 μm, classification precision index is 0.74, Newton classification efficiency is 74% and degree of dispersion is 1.27. This evaluation model avoids the limitation of evaluation for the single classification performance index and incomplete information got by the means of single factor experiment of operational parameters. It also provides the quantitative evaluating criteria for classification performance of a turbo air classifier, which offers a theoretical basis for effective production. This multi‐objective programming optimizing method and evaluation model on classification performance can be applied to other dynamic air classifiers as well.     

Diagnosis of a dermatological lesion using intelligent feature selection technique

The content-based image retrieval (CBIR) in dermatological diagnosis context, the information matching is the major concern in terms of feature vector-based classification. The discrimination of the feature vector leads to better classification as well as retrieval rate. Better retrieval results help the dermatologist to improve the diagnosis. In this paper, we proposed a support vector machine weight map (SVM W-Map)-based feature selection along with multi-class particle swarm optimization (PSO) presented for multi-class dermatological imaging dataset. The performance of the system was tested on a dataset including 1450 images and obtained 99.7% for specificity and 95.89% for sensitivity. The analysis and evaluations of results show that the proposed system has higher diagnosis ability when compared with other works.     

Earlier detection of cancer regions from MR image features and SVM classifiers

In this article, we examine the use of several segmentation algorithms for medical image classification. This work detects the cancer region from magnetic resonance (MR) images in earlier stage. This is accomplished in three stages. In first stage, four kinds of region‐based segmentation techniques are used such as K‐means clustering algorithm, expectation–maximization algorithm, partial swarm optimization algorithm, and fuzzy c‐means algorithm. In second stage, 18 texture features are extracting using gray level co‐occurrence matrix (GLCM). In stage three, classification is based on multi‐class support vector machine (SVM) classifier. Finally, the performance analysis of SVM classifier is analyzed using the four types of segmentation algorithm for a group of 200 patients (32—Glioma, 32—Meningioma, 44—Metastasis, 8—Astrocytoma, 72—Normal). The experimental results indicate that EM is an efficient segmentation method with 100% accuracy. In SVM, quadratic and RBF (σ = 0.5) kernel methods provide the highest classification accuracy compared to all other SVM kernel methods. © 2016 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 26, 196–208, 2016     

MRI brain image classification—a hybrid approach

The present article proposes a novel computer‐aided diagnosis (CAD) technique for the classification of the magnetic resonance brain images. The current method adopt color converted hybrid clustering segmentation algorithm with hybrid feature selection approach based on IGSFFS (Information gain and Sequential Forward Floating Search) and Multi‐Class Support Vector Machine (MC‐SVM) classifier technique to segregate the magnetic resonance brain images into three categories namely normal, benign and malignant. The proposed hybrid evolutionary segmentation algorithm which is the combination of WFF(weighted firefly) and K‐means algorithm called WFF‐K‐means and modified cuckoo search (MCS) and K‐means algorithm called MCS‐K‐means, which can find better cluster partition in brain tumor datasets and also overcome local optima problems in K‐means clustering algorithm. The experimental results show that the performance of the proposed algorithm is better than other algorithms such as PSO‐K‐means, color converted K‐means, FCM and other traditional approaches. The multiple feature set comprises color, texture and shape features derived from the segmented image. These features are then fed into a MC‐SVM classifier with hybrid feature selection algorithm, trained with data labeled by experts, enabling the detection of brain images at high accuracy levels. The performance of the method is evaluated using classification accuracy, sensitivity, specificity, and receiver operating characteristic (ROC) curves. The proposed method provides highest classification accuracy of greater than 98% with high sensitivity and specificity rates of greater than 95% for the proposed diagnostic model and this shows the promise of the approach. © 2015 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 25, 226–244, 2015     

Residual learning based CNN for breast cancer histopathological image classification

Biopsy is one of the most commonly used modality to identify breast cancer in women, where tissue is removed and studied by the pathologist under the microscope to look for abnormalities in tissue. This technique can be time-consuming, error-prone, and provides variable results depending on the expertise level of the pathologist. An automated and efficient approach not only aids in the diagnosis of breast cancer but also reduces human effort. In this paper, we develop an automated approach for the diagnosis of breast cancer tumors using histopathological images. In the proposed approach, we design a residual learning-based 152-layered convolutional neural network, named as ResHist for breast cancer histopathological image classification. ResHist model learns rich and discriminative features from the histopathological images and classifies histopathological images into benign and malignant classes. In addition, to enhance the performance of the developed model, we design a data augmentation technique, which is based on stain normalization, image patches generation, and affine transformation. The performance of the proposed approach is evaluated on publicly available BreaKHis dataset. The proposed ResHist model achieves an accuracy of 84.34% and an F1-score of 90.49% for the classification of histopathological images. Also, this approach achieves an accuracy of 92.52% and F1-score of 93.45% when data augmentation is employed. The proposed approach outperforms the existing methodologies in the classification of benign and malignant histopathological images. Furthermore, our experimental results demonstrate the superiority of our approach over the pre-trained networks, namely AlexNet, VGG16, VGG19, GoogleNet, Inception-v3, ResNet50, and ResNet152 for the classification of histopathological images.     

Content‐based gastric image retrieval using convolutional neural networks

The endoscopy procedure has demonstrated great efficiency in detecting stomach lesions, with extensive numbers of endoscope images produced globally each day. The content‐based gastric image retrieval (CBGIR) system has demonstrated substantial potential in gastric image analysis. Gastric precancerous diseases (GPD) have higher prevalence in gastric cancer patients. Thus, effective intervention is crucial at the GPD stage. In this paper, a CBGIR method is proposed using a modified ResNet‐18 to generate binary hash codes for a rapid and accurate image retrieval process. We tested several popular models (AlexNet, VGGNet and ResNet), with ResNet‐18 determined as the optimum option. Our proposed method was valued using a GPD data set, resulting in a classification accuracy of 96.21 ± 0.66% and a mean average precision of 0.927 ± 0.006 , outperforming other state‐of‐art conventional methods. Furthermore, we constructed a Gastric‐Map (GM) based on feature representations in order to visualize the retrieval results. This work has great auxiliary significance for endoscopists in terms of understanding the typical GPD characteristics and improving aided diagnosis.     

Magnetic resonance brain image classification based on weighted‐type fractional Fourier transform and nonparallel support vector machine

To classify brain images into pathological or healthy is a key pre‐clinical state for patients. Manual classification is tiresome, expensive, time‐consuming, and irreproducible. In this study, we aimed to present an automatic computer‐aided system for brain‐image classification. We used 90 T2‐weighted images obtained by magnetic resonance images. First, we used weighted‐type fractional Fourier transform (WFRFT) to extract spectrums from each magnetic resonance image. Second, we used principal component analysis (PCA) to reduce spectrum features to only 26. Third, those reduced spectral features of different samples were combined and were fed into support vector machine (SVM) and its two variants: generalized eigenvalue proximal SVM and twin SVM. A 5 × 5‐fold cross‐validation results showed that this proposed “WFRFT + PCA + generalized eigenvalue proximal SVM” yielded sensitivity of 99.53%, specificity of 92.00%, precision of 99.53%, and accuracy of 99.11%, which are comparable with the proposed “WFRFT + PCA + twin SVM” and better than the proposed “WFRFT + PCA + SVM.” Besides, all three proposed methods were superior to eight state‐of‐the‐art algorithms. Thus, WFRFT is effective, and the proposed methods can be used in practical. © 2015 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 25, 317–327, 2015     

基于SVM的高通量dPCR基因芯片荧光图像分类研究

目的 为了实现高通量dPCR基因芯片荧光图像的亮点分类与计数,提出一种基于支持向量机（SVM）的荧光图像分类与计数方法。方法 首先对荧光图像进行去噪、对比度增强等图像预处理,对预处理后荧光图像进行亮点区域提取标注,去除背景与暗点的冗余信息,利用方向梯度直方图（Histogram of Oriented Gradient, HOG）提取鉴别特征,计算合并所有样本的亮点特征得到HOG特征向量,根据已得到的HOG特征向量创建一个线性SVM分类器,利用训练好的SVM分类器对荧光图像亮点进行分类与计数。结果 对比传统算法,文中算法具有较高的分类识别精度,平均准确率高达98%以上,可以很好地实现荧光图像亮点分类与计数。结论 在有限的小样本标注数据下,文中算法具有良好的分类性能,能够有效识别荧光图像中的亮点,对其他荧光图像分类研究也具有一定参考价值。     

Learning methods for melanoma recognition

Melanoma is the most deadly skin cancer. Early diagnosis is a challenge for clinicians. Current algorithms for skin lesions' classification focus mostly on segmentation and feature extraction. This article instead puts the emphasis on the learning process, testing the recognition performance of three different classifiers: support vector machine (SVM), artificial neural network and k‐nearest neighbor. Extensive experiments were run on a database of more than 5000 dermoscopy images. The obtained results show that the SVM approach outperforms the other methods reaching an average recognition rate of 82.5% comparable with those obtained by skilled clinicians. If confirmed, our data suggest that this method may improve classification results of a computer‐assisted diagnosis of melanoma. © 2010 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 20, 316–322, 2010     

Defective kernel detection using a linear colour CCD

Abstract

This study was aimed at detecting defective wheat (Triticum durum Desf) with a machine vision system of linear colour charge-coupled device. One thousand one hundred and sixty-nine images were captured for sound kernels, 710 for black germ kernels and 627 for broken kernels. A software package was developed to extract various morphological, colour and texture features from the images captured. Then the experimental data were subjected to multivariate analysis. Principal component analysis was employed to differentiate samples from different categories. Partial least square discriminant analysis and venetian blinds cross-validation were used to develop classification models. The best detection accuracies of samples were 92·7, 88·0 and 89·6% for black germ kernels, broken kernels and sound kernels. The results have proved that it is feasible and effective to employ partial least square discriminant analysis for feature selection and defective kernel detection.     

改进MDSMOTE与PSO-SVM在汽车组合仪表分类预测中的应用

汽车组合仪表生产过程中质检项目多且检测时间长,这在一定程度上制约了其生产效率的进一步提升。为此,提出一种基于改进最远点合成少数类过采样技术（max distance synthetic minority over-sampling technique,MDSMOTE）的支持向量机（support vector machine, SVM）分类预测方法。首先,结合专家经验对汽车组合仪表的原始生产数据进行特征筛选,并在MDSMOTE中引入类不平衡率I_R,以对所筛选的特征数据进行扩充;然后,利用粒子群优化（particle swarm optimization, PSO）算法对SVM的误差惩罚因子C和核函数参数γ进行优化;最后,建立优化的SVM分类预测模型,并对汽车组合仪表进行分类。通过与其他分类预测模型在不同数据集上的预测结果进行对比可知,基于改进MDSMOTE的SVM分类预测模型的准确率、F值和几何平均值等评价指标均优于其他模型。所提出方法在汽车仪表产品分类上表现出较强的泛化能力和稳定性,可为仪表制造企业生产效率的提升提供有效参考。     

Single-crystal neutron diffraction study of ferroelectric Ba0·92Ca0·08TiO3

Crystal structure of BaTiO₃ doped with 8% Ca²⁺ is refined using single-crystal neutron diffraction data and it is shown that the doped Ca²⁺ ion substitutes only at the Ba sites. The refined cell (P4 mm) parameters area=b=3·982(3) Å,c=4·003(3) Å with a finalR value of 0·02 (onF). Existence of multiple domains in the crystal is ruled out based on refinement with multidomain model.     

Analysis of constitutive behaviours and processing map of 7020 aluminium alloy

Abstract

The authors present a study on the hot formability of 7020 aluminium alloy. Isothermal hot compression tests of solid cylindrical specimens were performed in the temperature range of 300–550°C and the strain rate range of 0·001–10 s^–1. Stress–strain curves obtained from the experiment data are fitted using the Sellars–Tegart constitutive equation to obtain the constitutive parameters. Using the dynamic material model, the authors develop a processing map based on the flow stress data. The map shows that the parameters suitable for hot working are a temperature range of 450–550°C and a strain rate range of 0·001–0·1 s^–1. This parameter range is where the efficiency of power dissipation is above 27% and where dynamic recrystallisation occurs. Unstable regions to be avoided in hot forming are deduced from an instability condition. The processing map is validated by comparing the microstructures of deformed compression specimens.     

An approach for classification of malignant and benign microcalcification clusters

The only reliable and successful treatment of breast cancer is its detection through mammography at initial stage. Clusters of microcalcifications are important signs of breast cancer. Manual interpretation of mammographic images, in which the suspicious regions are indicated as areas of varying intensities, is not error free due to a number of reasons. These errors can be reduced by using computer-aided diagnosis systems that result in reduction of either false positives or true negatives. The purpose of the study in this paper is to develop a methodology for distinguishing malignant microcalcification clusters from benign microcalcification clusters. The proposed approach first enhances the region of interest by using morphological operations. Then, two types of features, cluster shape features and cluster texture features, are extracted. A Support Vector Machine is used for classification. A new set of shape features based on the recursive subsampling method is added to the feature set, which improves the classification accuracy of the system. It has been found that these features are capable of differentiating malignant and benign tissue regions. To investigate the performance of the proposed approach, mammogram images are taken from Digital Database for Screening Mammography database and an accuracy of 94.25% has been achieved. The experiments have shown that the proposed classification system minimizes the classification errors and is more efficient in correct diagnosis.     

Pt–Ru decorated self-assembled TiO2–carbon hybrid nanostructure for enhanced methanol electrooxidation

Porous titanium oxide–carbon hybrid nanostructure (TiO₂–C) with a specific surface area of 350 m²/g and an average pore-radius of 21?·?8 Å is synthesized via supramolecular self-assembly with an in situ crystallization process. Subsequently, TiO₂–C supported Pt–Ru electro-catalyst (Pt–Ru/TiO₂–C) is obtained and investigated as an anode catalyst for direct methanol fuel cells (DMFCs). X-ray diffraction, Raman spectroscopy and transmission electron microscopy (TEM) have been employed to evaluate the crystalline nature and the structural properties of TiO₂–C. TEM images reveal uniform distribution of Pt–Ru nanoparticles (d _Pt???Ru ?=?1·5–3·5 nm) on TiO₂–C. Methanol oxidation and accelerated durability studies on Pt–Ru/TiO₂–C exhibit enhanced catalytic activity and durability compared to carbon-supported Pt–Ru. DMFC employing Pt–Ru/TiO₂–C as an anode catalyst delivers a peak-power density of 91 mW/cm² at 65 °C as compared to the peak-power density of 60 mW/cm² obtained for the DMFC with carbon-supported Pt–Ru anode catalyst operating under similar conditions.     

An intelligent mining system for diagnosing medical images using combined texture‐histogram features

The aim of this article is to design an expert system for medical image diagnosis. We propose a method based on association rule mining combined with classification technique to enhance the diagnosis of medical images. This system classifies the images into two categories namely benign and malignant. In the proposed work, association rules are extracted for the selected features using an algorithm called AprioriTidImage, which is an improved version of Apriori algorithm. Then, a new associative classifier CLASS_Hiconst ( CL assifier based on ASS ociation rules with Hi gh Con fidence and S uppor t ) is modeled and used to diagnose the medical images. The performance of our approach is compared with two different classifiers Fuzzy‐SVM and multilayer back propagation neural network (MLPNN) in terms of classifier efficiency with sensitivity, specificity, accuracy, positive predictive value, and negative predictive value. The experimental result shows 96% accuracy, 97% sensitivity, and 96% specificity and proves that association rule based classifier is a powerful tool in assisting the diagnosing process. © 2013 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 23, 194–203, 2013     

Analysis and optimization of process parameters affecting classification performances indices of the turbo air classifier

Classification performance indices of a turbo air classifier mainly include cut size, classification precision, and Newton's classification efficiency, which would be effected by the process parameters including air inlet velocity, rotor cage's rotary speed and feeding speed. Orthogonal experiment method was used to analyze the influence of process parameters on the classification performance indices according to quartz sand powder classification experimental data as well as the calcium carbonate classification experimental data. Through range analysis of the orthogonal experiments for quartz sand and calcium carbonate, it is found that air inlet velocity and rotor cage's rotary speed play an important role on classification performance indices including classification accuracy, cut size and its corresponding Newton's classification efficiency. Feeding speed has slight impact on these performance indices. The best optimized combination of the process parameters for quartz sand was: when air inlet velocity was 20 m · s^–1, rotor cage rotary speed was 900 min^–1 and feeding speed was 56.69 kg · h^–1, the high Newton's classification efficiency and classification accuracy could be obtained and the corresponding values were 73% and 0.66. This best optimized combination was among the 9 group orthogonal experiments for quartz sand. The best optimized combination of the process parameters for calcium carbonate was: air inlet velocity was 14 m · s^–1, rotor cage rotary speed was 1200 min^–1 and feeding speed was 120 kg · h^–1. However this best optimized combination wasn't among the 9 group orthogonal experiments for calcium carbonate. The evaluated value of the Newton's classification efficiency of cut size can be obtained, which was 73%. In order to verify this evaluated value, the classification experiment for calcium carbonate with this best optimized combination of process parameters was carried out and experimental value of Newton's classification efficiency was 75%, which was close to the evaluated value. Through orthogonal experiment analysis, the influence law of process parameters on classification performances indices of the turbo air classifier can be acquired and the optimized combination of process parameters can be obtained. It paves the way of the development of the turbo air classifier.