Spatial Gray Level Dependence Matrix Research Articles

A novel approach to brain tumor detection using K-Means++, SGLDM, ResNet50, and synthetic data augmentation.

Introduction: Brain tumors are abnormal cell growths in the brain, posing significant treatment challenges. Accurate early detection using non-invasive methods is crucial for effective treatment. This research focuses on improving the early detection of brain tumors in MRI images through advanced deep-learning techniques. The primary goal is to identify the most effective deep-learning model for classifying brain tumors from MRI data, enhancing diagnostic accuracy and reliability. Methods: The proposed method for brain tumor classification integrates segmentation using K-means++, feature extraction from the Spatial Gray Level Dependence Matrix (SGLDM), and classification with ResNet50, along with synthetic data augmentation to enhance model robustness. Segmentation isolates tumor regions, while SGLDM captures critical texture information. The ResNet50 model then classifies the tumors accurately. To further improve the interpretability of the classification results, Grad-CAM is employed, providing visual explanations by highlighting influential regions in the MRI images. Result: In terms of accuracy, sensitivity, and specificity, the evaluation on the Br35H::BrainTumorDetection2020 dataset showed superior performance of the suggested method compared to existing state-of-the-art approaches. This indicates its effectiveness in achieving higher precision in identifying and classifying brain tumors from MRI data, showcasing advancements in diagnostic reliability and efficacy. Discussion: The superior performance of the suggested method indicates its robustness in accurately classifying brain tumors from MRI images, achieving higher accuracy, sensitivity, and specificity compared to existing methods. The method's enhanced sensitivity ensures a greater detection rate of true positive cases, while its improved specificity reduces false positives, thereby optimizing clinical decision-making and patient care in neuro-oncology.

An Effective approach for Plant Disease Detection Using Assessment-Based Convolutional Neural Networks (A-CNN)

Agriculture is crucial in determining land availability and economic productivity in developing countries, where most of the population relies on this sector. Even though plant diseases are common, they are primarily identified in agriculture. Automated disease detection technology is necessary for the timely identification of plant diseases. Detecting plant diseases is crucial in avoiding loss of production and enhancing agricultural produce quality. Lack of proper attention in this area can cause severe damage to plants, resulting in loss of product quality, quantity, or productivity. Diagnosing plant diseases can be a complex process that requires specialized knowledge and hands-on attention. The previous methods didn't find accurate disease in the plant leaves. To combat this problem, this paper presents deep learning (DL) based Assessment-based Convolutional Neural Network (A-CNN) method to detect healthy and unhealthy plant leaves. We first collect the plant image dataset from a Kaggle repository. To enhance the quality of plant images and achieve a smoother appearance, an Optimized Gaussian Wiener Filter (OGWF) can be utilized for image pre-processing. Additionally, the edges of plant leaves can be detected by implementing Sobel and Canny operators. Then we use Otsu's Threshold Fragment (OTF) algorithm to segment the disease-affected region. Additionally, the Spatial Grey-Level Dependence Matrix (SGLDM) algorithm is utilized to identify the most suitable feature of the impacted leaf. The A-CNN method is employed to detect healthy and unhealthy plant leaves accurately. The proposed simulation results have shown higher accuracy, sensitivity, and specificity in predicting plant diseases than other methods.

Osteosarcoma detection in histopathology images using ensemble machine learning techniques

Osteosarcoma is a bone malignancy that severely affects the long bones of a human arm or leg. Histopathological analysis with ML (Machine Learning) algorithm offers a gainful way to examine the difference in the osteosarcoma texture. This work presents a new ensemble learning (EL) model with severity analysis in classifying bone tumor (cancer). Initially, pre-processing includes noise removal with Weighted Bilateral (WB) filtering, image quality enhancement using Parabolic Balance Contrast Enhancement (PBCE), and unwanted regions are discarded using erosion and dilation operations. The SIKC (Superpixel Improved K-means Clustering) approach characterizes the segmentation process. The extraction of valuable information is done with color and texture features, namely RGB histogram and Spatial Gray Level Dependence Matrix (Spatial GLDM). Next, the appropriate features are selected using Max_Relevance Min_Redundancy (MRMR) technique. Finally, classification is performed using EL models such as Kernel-Support Vector Machine (KSVM), Improved ANN with Beetle Swarm Optimization (IANN_BSO), and Kernel Extreme Learning Machine Optimized with Chaotic Salp Swarm (OKELM), which categorizes the images into Viable, Non-tumor and Non-viable type tumor along with severity analysis. The proposed study used a PYTHON tool for implementation, and the simulation is done by utilizing publicly accessible histopathological images obtained from TCIA (The Cancer Imaging Archive). The performance of a proposed EL model is evaluated using accuracy, specificity, sensitivity, kappa, F-measure, ROC (AUC) etc. Thus, the proposed EL model achieved the highest results in terms of accuracy (98.505%) compared to other classifiers such as AdaBoost (91.70%), ELM (96.41%), KNN (86.90%), SVM (83.80%), and ANN (90.13%) respectively.

Automatic Multi-Class Brain Tumor Classification Using Residual Network-152 Based Deep Convolutional Neural Network

Brain tumor is one of the leading causes of death in humans worldwide. Image recognition or computer vision uses deep learning based approaches for automatic tumor detection by classifying brain images. It is difficult to analyze the similarity between brain tissues while processing the magnetic resonance imaging (MRI) brain images for tumor classification. In this paper, residual network-152 (ResNet-152) with softmax layer is proposed for accurate detection of brain tumor with low complexity. Initially, the brain images are pre-processed and segmented with adaptive canny mayfly algorithm (ACMA). More discriminative features are extracted from the pre-processed image with spatial gray level dependence matrix (SGLDM), and optimal features are selected with modified chimpanzee optimization algorithm (MChOA). The optimal feature selection and optimal performance of classification are obtained by eliminating poor generalization and over specialization. After eliminating redundancies, the features are fed to residual classification. The overall performance of the proposed tumor classification method is evaluated using various parameters such as accuracy, precision, recall, F-score, MCC and balanced accuracy. The evaluation results indicate that our proposed method reached the accuracy level of 98.85%, which is efficient than other conventional approaches such as convolutional neural network (CNN), ResNet, recurrent neural network (RNN), random belief network (RBN), liner support vector machine (LSVM) and poly-SVM.

Optimized U‐Net convolutional neural network based breast cancer prediction for accuracy increment in big data

SummaryBig data is data collected with huge dimensions and continuous exponential growth over time. In recent years, big data in health care has been commonly used to predict diseases. Breast cancer is one of the most common diseases and the secondary cause of death among women. Early diagnosis of breast cancer can prevent the risk of death. Few types of research have been done on breast cancer prediction on big data. However, the traditional prediction models have less efficient in terms of accuracy and error rate. The Optimized U‐Net Convolutional neural network (OU‐NetCNN) model is proposed in this paper to overcome these challenges. Hadoop is the storage system generated to store the datasets samples for big data. The data samples from two datasets, namely BreakHis and Kaggle (Breast Histopathology Images), are preserved in this storage system. The BreakHis data are considered for further processes like pre‐processing, segmentation, feature extraction, feature selection, and classification from the stored data samples. The noise is removed from the histopathological breast images in pre‐processing using the adaptive fast peer‐group filtering (AFPGF) approach. Then the morphological operations such as erosion and dilation are used to eradicate unwanted portions and the quality of the image is enhanced using the improved balance contrast enhancement method (IBCE). Next, edges of breast images are detected using an adaptive artificial ecosystem optimization (AAEO) algorithm‐based edge detection approach in the segmentation process. The features are extracted using the Spatial Gray Level Dependence Matrix (SGLDM) and the optimal features are selected by the modified selfish herd optimization (MSHO) algorithm. Finally, the selected features are fed into the proposed OU‐NetCNN model to classify the histopathology images as benign and malignant images. This hybridization minimizes the error rate, computational complexity and over fitting issues. The simulation analysis is performed in the PYTHON tool. Two datasets, namely BreakHis and Kaggle (Breast Histopathology Images), are considered. Some of the measures such as precision, sensitivity, accuracy and F‐measure are considered to evaluate the performance of the proposed model and compared with existing approaches.

Elucidation of microstructural changes in leaves during senescence using spectral domain optical coherence tomography

Leaf senescence provides a unique window to explore the age-dependent programmed degradation at organ label in plants. Here, spectral domain optical coherence tomography (SD-OCT) has been used to study in vivo senescing leaf microstructural changes in the deciduous plant Acer serrulatum Hayata. Hayata leaves show autumn phenology and change color from green to yellow and finally red. SD-OCT image analysis shows distinctive features among different layers of the leaves; merging of upper epidermis and palisade layers form thicker layers in red leaves compared to green leaves. Moreover, A-scan analysis showed a significant (p < 0.001) decrease in the attenuation coefficient (for wavelength range: 1100–1550 nm) from green to red leaves. In addition, the B-scan analysis also showed significant changes in 14 texture parameters extracted from second-order spatial gray level dependence matrix (SGLDM). Among these parameters, a set of three features (energy, skewness, and sum variance), capable of quantitatively distinguishing difference in the microstructures of three different colored leaves, has been identified. Furthermore, classification based on k-nearest neighbors algorithm (k-NN) was found to yield 98% sensitivity, 99% specificity, and 95.5% accuracy. Following the proposed technique, a portable noninvasive tool for quality control in crop management can be anticipated.

&lt;i&gt;Indic&lt;/i&gt; script identification from handwritten document images

Script identification plays an important role in document image processing especially for multilingual environment. This paper hires two conventional textural methods for recognition of the scripts of the handwritten documents inscribed in different Indic scripts. The first method extracts well-known Haralick features from spatial grey-level dependence matrix (SGLDM) and the second method computes fractal dimension by using segmentation-based fractal texture analysis (SFTA). Finally, a 104-element feature vector is constructed from each page image by these two methods. The proposed technique is then evaluated on a total dataset comprising 360 handwritten document pages written in 12 Indian official scripts namely Bangla, Devanagari, Gujarati, Gurumukhi, Kannada, Malayalam, Manipuri, Oriya, Tamil, Telugu, Urdu and Roman. Experimentations using multiple classifiers reveal that multilayer perceptron (MLP) shows the highest identification accuracy of 96.94%. The encouraging outcome confirms the efficacy of customary textural features to handwritten Indic script identification.

Feature Set Fusion for Spoof Iris Detection

Iris recognition is considered as one of the most promising noninvasive biometric systems providing automated human identification. Numerous programs, like unique ID program in India - Aadhar, include iris biometric to provide distinctive identity identification to citizens. The active area is usually captured under non ideal imaging conditions. It usually suffers from poor brightness, low contrast, blur due to camera or subject's relative movement and eyelid eyelash occlusions. Besides the technical challenges, iris recognition started facing sophisticated threats like spoof attacks. Therefore it is vital that the integrity of such large scale iris deployments must be preserved. This paper presents the development of a new spoof resistant approach which exploits the statistical dependencies of both general eye and localized iris regions in textural domain using spatial gray level dependence matrix (SGLDM), gray level run length matrix (GLRLM) and contourlets in transform domain. We did experiments on publicly available fake and lens iris image databases. Correct classification rate obtained with ATVS-FIr iris database is 100% while it is 95.63% and 88.83% with IITD spoof iris databases respectively.

CAROTID WALL MEASUREMENT AND ASSESSMENT BASED ON PIXEL-BASED AND LOCAL TEXTURE DESCRIPTORS

Aim of this paper is to develop an automated system for the classification and characterization of carotid wall status and to develop a robust system based on local texture descriptors. A database of 200 longitudinal ultrasound images of carotid artery is used. One-hundred images with Intima-Media Thickness (IMT) value higher than 0.8[Formula: see text]mm are considered as high risk. Six different rectangular pixel neighborhoods were considered: four areas centered on the selected element, with sizes [Formula: see text], [Formula: see text], [Formula: see text], and [Formula: see text] pixels, and two noncentered areas with sizes [Formula: see text] pixels upwards and downwards. We have extracted various texture descriptors (31 based on the co-occurrence gray level matrix, 13 based on the spatial gray level dependence matrix, and 20 based on the gray level run length matrix (GLRLM) from neighborhood. We have used Quick Reduct Algorithm to select 12 most discriminant features from extracted 211 features. Each pixel is then assigned to the vessel lumen, to the intima-media complex, or to the adventitia by using an integrated system of three feed-forward neural networks. The boundaries between the three regions are used to estimate the IMT value. The texture features associated with GLRLM are found to be clinically most significant. We have obtained an overall classification accuracy of 79.5%, sensitivity of 87%, and specificity of 72%. We observed a unique classification pattern between low risk and high risk images: in the latter ones, a considerable number of pixels of the intima–media complex ([Formula: see text]) was classified as belonging to the adventitia. This percentage is statistically higher than that of low risk images ([Formula: see text]; [Formula: see text]). Locally extracted and pixel-based descriptors are able to capture the inner characteristics of the carotid wall. The presence of misclassified pixels in the intima–media complex is associated to higher cardiovascular risk.

Classification of degradation in oil-impregnated cellulose insulation using texture analysis of optical microscopy images

Oil-impregnated Kraft paper samples normally used for power transformer electrical insulation have been artificially aged thermally to produce a sample set with varying levels of insulation deterioration. The samples were aged in an oven at a temperature above the normal operating temperature of a power transformer.Digital images obtained from optical microscopy measurements on the paper samples have been analyzed using a texture analysis method which converts each image into a spatial gray level dependence matrix (SGLDM). The SGLDM contains information about the statistical variation of pixels gray-level intensities in an image and thereby information about the sample texture. Mathematical operators applied to the SGLDM have been used to extract 22 different statistical texture features for each sample image.Optical microscopy images of the thermally aged samples show that thermal deterioration of the insulation paper produces changes in morphology and physical structure. These changes are detectable by the statistical texture features extracted from the SGLDM texture analysis. Statistical classification is performed on the feature set to demonstrate that differentiation between oil-impregnated paper samples with different levels of thermal degradation is reliable with low error rates. Therefore, development of a practical method to assess condition of oil-impregnated paper insulation using optical microscopy and texture analysis is promising.

Surface Roughness Detection of Arteries via Texture Analysis of Ultrasound Images for Early Diagnosis of Atherosclerosis

There is a strong research interest in identifying the surface roughness of the carotid arterial inner wall via texture analysis for early diagnosis of atherosclerosis. The purpose of this study is to assess the efficacy of texture analysis methods for identifying arterial roughness in the early stage of atherosclerosis. Ultrasound images of common carotid arteries of 15 normal mice fed a normal diet and 28 apoE−/− mice fed a high-fat diet were recorded by a high-frequency ultrasound system (Vevo 2100, frequency: 40 MHz). Six different texture feature sets were extracted based on the following methods: first-order statistics, fractal dimension texture analysis, spatial gray level dependence matrix, gray level difference statistics, the neighborhood gray tone difference matrix, and the statistical feature matrix. Statistical analysis indicates that 11 of 19 texture features can be used to distinguish between normal and abnormal groups (p<0.05). When the 11 optimal features were used as inputs to a support vector machine classifier, we achieved over 89% accuracy, 87% sensitivity and 93% specificity. The accuracy, sensitivity and specificity for the k-nearest neighbor classifier were 73%, 75% and 70%, respectively. The results show that it is feasible to identify arterial surface roughness based on texture features extracted from ultrasound images of the carotid arterial wall. This method is shown to be useful for early detection and diagnosis of atherosclerosis.

Detection of Breast Cancer Cells by Using Texture Analysis

In Brief Breast cancer is a serious public health problem worldwide. This study presents a texture analysis approach to detect cancer on mammograms. Fifty-five regions of interest (ROIs) containing abnormal breast tissue and 192 ROIs containing normal breast tissue were extracted from digital mammograms obtained from the Mammographic Image Analysis Society database. Haralick texture features derived from spatial grey-level dependence matrix were calculated for each ROI. The importance of each feature in distinguishing abnormal from normal tissues was determined by stepwise linear discriminant analysis. The methodology obtained 92.7% of accuracy for the detection of abnormal tissue (cancer) on digitalized mammograms. Breast cancer is a serious public health problem worldwide. This study presents a texture analysis approach to detect cancer on mammograms.

Automatic classification of brain computed tomography images using wavelet-based statistical texture features

Automated and accurate classification of computed tomography (CT) images is an integral component of the analysis and interpretation of neuro imaging. In this paper, we present the wavelet-based statistical texture analysis method for the classification of brain tissues into normal, benign, malignant tumor of CT images. Comparative studies of texture analysis method are performed for the proposed texture analysis method and spatial gray level dependence matrix method (SGLDM). Our proposed system consists of five phases (i) image acquisition, (ii) discrete wavelet decomposition (DWT), (iii) feature extraction, (iv) feature selection, and (v) analysis of extracted texture features by classifier. A wavelet-based statistical texture feature set is derived from two level discrete wavelet transformed approximation (low frequency part of the image) sub image. Genetic algorithm (GA) and principal component analysis (PCA) are used to select the optimal texture features from the set of extracted features. The support vector machine (SVM) is employed as a classifier. The results of SVM for the texture analysis methods are evaluated using statistical analysis and receiver operating characteristic (ROC) analysis. The experimental results show that the proposed system is able to achieve higher classification accuracy effectiveness as measured by sensitivity and specificity.

Quantitative analysis of colorectal lesions observed on magnified endoscopy images

Various surface mucosal pit patterns, as recognized by endoscopists, correlate with the histologic features of colorectal cancers. We investigated whether magnified endoscopy images of these pit patterns could be analyzed quantitatively and thus facilitate computer-aided diagnosis of colorectal lesions. We applied both texture analysis and scale-invariant feature transform (SIFT) descriptors and discriminant analysis to magnified endoscopy images of 165 neoplastic colorectal lesions (pit patterns: type III(L)/IV, n = 44; type V(I)-mildly irregular, n = 36; type V(I)-severely irregular, n = 45; type V(N), n = 40) [histologic findings: tubular adenoma (TA), n = 56; carcinoma with intramucosal or even scant submucosal invasion (M/SM-s), n = 52, carcinoma with massive submucosal invasion (SM-m), n = 57]. We analyzed differences in pit pattern values and corresponding histologic values to determine whether the values were diagnostically meaningful. Gray-level difference matrix (GLDM) inverse difference moment and spatial gray-level dependence matrix (SGLDM) local homogeneity values differed significantly between type III(L)/IV and type V(N) pit patterns. Values differed significantly for each analyzed feature between type III(L)/IV and type V(I)-severely irregular patterns and were high but descending for type III(L)/IV, type V(I)-mildly irregular, and type V(I)-severely irregular pit patterns (in that order). Similarly, texture analysis yielded high but descending values for TA, M/SM-s, and SM-m (in that order). Furthermore, SIFT descriptors and discriminant analysis yielded differences that were superior to those obtained by texture analyses. Computer analysis of magnified endoscopy images for the diagnosis of colorectal lesions appears feasible. We anticipate further developments in the computer-aided diagnosis of pit patterns on magnified endoscopy images.

Biofilm monitoring on rotating discs by image analysis

The macrostructure development of biofilms grown in a lab-scale rotating biological contactor was monitored by analyzing the average opacity and the texture of gray-level images of the discs. The reactor was fed with municipal or synthetic wastewater. Experiments lasted on average 4-14 weeks. The images were obtained with a flat-bed scanner. The opacity and its standard deviation are directly extracted from the annular zone where the biofilm develops. This zone is defined by the outer edge of the disc and the waterline. The spatial gray-level dependence matrix (SGLDM) approach was used for the texture assessment. As this method requires rectangular images, a geometrical transformation had to be developed to transform the ring into a workable area. This transformation now allows quantitative image analysis on circular biofilms. As a last step, Principal Components Analysis was applied to the set of textural descriptors to reduce the number of textural parameters. Opacity and textural information allowed the non-intrusive monitoring of the growth/regrowth of the biofilms as well as biofilm loss, due to detachment, auto-digestion, or protozoan grazing. Textural description was very valuable by helping to discriminate biofilms of similar opacity characteristics but presenting different macrostructures.

Analyzing characteristic length scales in biofilm structures

The quantification of biofilm structure based on image analysis requires a statistical measure like representative elemental areas (REA) to determine the necessary size of biofilm area to be imaged. In this study, REAs for biofilm structure were calculated for the descriptors Gray level and Correlation (COR) derived from a spatial gray level dependence matrix analysis (SGLDM). An important difference between these two descriptors is their response to structural features at different spatial scales. Gray level is a scale-independent descriptor, whereas COR is scale-dependent. For scale-independent descriptors, the size of the individual images is not relevant when determining REAs. This is in contrast to scale-dependent descriptors for which REAs can only be determined when the area of each image covers the range of structural variability of the biofilm. We used COR to analyze scale dependence of structural heterogeneity at different length scales. A characteristic length of 400 microm in biofilm images provides structural information relevant for mass transport phenomena in biofilms. Overall REAs for gray level and COR were on average 3.4 mm(2). The scale-dependent descriptor COR could not in all cases accurately be determined from combining individual image analysis results--even when the combined area resulted in the REA. Microscope and camera specifications define the upper and lower limit of detectable characteristic length that can be extracted from images and should therefore be considered in the experimental design when choosing an imaging technique.

Differential diagnosis of CT focal liver lesions using texture features, feature selection and ensemble driven classifiers

The aim of the present study is to define an optimally performing computer-aided diagnosis (CAD) architecture for the classification of liver tissue from non-enhanced computed tomography (CT) images into normal liver (C1), hepatic cyst (C2), hemangioma (C3), and hepatocellular carcinoma (C4). To this end, various CAD architectures, based on texture features and ensembles of classifiers (ECs), are comparatively assessed. Number of regions of interests (ROIs) corresponding to C1-C4 have been defined by experienced radiologists in non-enhanced liver CT images. For each ROI, five distinct sets of texture features were extracted using first order statistics, spatial gray level dependence matrix, gray level difference method, Laws' texture energy measures, and fractal dimension measurements. Two different ECs were constructed and compared. The first one consists of five multilayer perceptron neural networks (NNs), each using as input one of the computed texture feature sets or its reduced version after genetic algorithm-based feature selection. The second EC comprised five different primary classifiers, namely one multilayer perceptron NN, one probabilistic NN, and three k-nearest neighbor classifiers, each fed with the combination of the five texture feature sets or their reduced versions. The final decision of each EC was extracted by using appropriate voting schemes, while bootstrap re-sampling was utilized in order to estimate the generalization ability of the CAD architectures based on the available relatively small-sized data set. The best mean classification accuracy (84.96%) is achieved by the second EC using a fused feature set, and the weighted voting scheme. The fused feature set was obtained after appropriate feature selection applied to specific subsets of the original feature set. The comparative assessment of the various CAD architectures shows that combining three types of classifiers with a voting scheme, fed with identical feature sets obtained after appropriate feature selection and fusion, may result in an accurate system able to assist differential diagnosis of focal liver lesions from non-enhanced CT images.

Combined Statistical and Multiscale View on Ultrasonic Liver Images for Characterization

Combined Statistical and Multiscale View on Ultrasonic Liver Images for Characterization

Usefulness of Texture Analysis for Computerized Classification of Breast Lesions on Mammograms

This work presents the usefulness of texture features in the classification of breast lesions in 5,518 images of regions of interest, which were obtained from the Digital Database for Screening Mammography that included microcalcifications, masses, and normal cases. Sixteen texture features were used, i.e., 13 were based on the spatial gray-level dependence matrix and 3 on the wavelet transform. The nonparametric K-NN classifier was used in the classification stage. The results obtained from receiver operating characteristic analysis indicated that the texture features can be used for separating normal regions and lesions with masses and microcalcifications, yielding the area under the curve (AUC) values of 0.957 and 0.859, respectively. However, the texture features were not very effective for distinguishing between malignant and benign lesions because the AUC was 0.617 for masses and 0.607 for microcalcifications. The study showed that the texture features can be used for the detection of suspicious regions in mammograms.

Characterisation of grinding media wear by visual texture

Three methods of automated characterisation of visual texture—spatial grey-level dependence matrix (SGLDM), grey-level run length (GLRL) and grey-level partial derivatives (GLPD)—have been applied to assess indirectly the wear of grinding media from series of grey-level photographs obtained by scanning electron microscopy (SEM) at different magnifications (between 350× and 5000×). The three methods are complementary, but for a routine analysis, the use of a single intermediate magnification (≈1500×), combined with the SGLDM method, could be advocated to compare the degree of wear.