Gray Level Co-occurrence Matrix Features Research Articles

Efficient optimization based thresholding technique for analysis of alzheimer MRIs

Purpose study: Alzheimer is a type of dementia that usually affects older adults by creating memory loss due to damaged brain cells. The damaged brain cells lead to shrinkage in the size of the brain and it is very difficult to extract the grey matter (GM) and white matter (WM). The segmentation of GM and WM is a challenging task due to its homogeneous nature between the neighborhood tissues. In this proposed system, an attempt has been made to extract GM and WM tissues using optimization-based segmentation techniques. Materials and methods: The optimization method is considered for the classification of normal and alzheimer disease (ad) through magnetic resonance images (mri) using a modified cuckoo search algorithm. Gray Level Co-Occurrence Matrix (GLCM) features are calculated from the extracted GM and WM. Principal Component Analysis (PCA) is adopted for selecting the best features from the GLCM features. Support Vector Machine (SVM) is a classifier which is used to classify the normal and abnormal images. Results: The proposed optimization algorithm provides most promising and efficient level of image segmentation compared to fuzzy c means (fcm), otsu, particle swarm optimization (pso) and cuckoo search (cs). The modified cuckoo yields high accuracy of 96%, sensitivity of 97% and specificity of 94% than other methods due to its powerful searching potential for the proper identification of gray and WM tissues. Conclusions: The results of the classification process proved the effectiveness of the proposed technique in identifying alzheimer affected patients due to its very strong optimization ability. The proposed pipeline helps to diagnose early detection of AD and better assessment of the neuroprotective effect of a therapy.

Classifying Lithofacies from Textural Features in Whole Core CT-Scan Images

Summary X-ray computerized tomography (CT) is a nondestructive method of providing information about the internal composition and structure of whole core reservoir samples. In this study we propose a method to classify lithology. The novelty of this method is that it uses statistical and textural information extracted from whole core CT images in a supervised learning environment. In the proposed approaches, first-order statistical features and textural grey-level co-occurrence matrix (GLCM) features are extracted from whole core CT images. Here, two workflows are considered. In the first workflow, the extracted features are used to train a support vector machine (SVM) to classify lithofacies. In the second workflow, a principal component analysis (PCA) step is added before training with two purposes: first, to eliminate collinearity among the features and second, to investigate the amount of information needed to differentiate the analyzed images. Before extracting the statistical features, the images are preprocessed and decomposed using Haar mother wavelet decomposition schemes to enhance the texture and to acquire a set of detail images that are then used to compute the statistical features. The training data set includes lithological information obtained from core description. The approach is validated using the trained SVM and hybrid (PCA + SVM) classifiers to predict lithofacies in a set of unseen data. The obtained results show that the SVM classifier can predict some of the lithofacies with high accuracy (up to 91% recall), but it misclassifies, to some extent, similar lithofacies with similar grain size, texture, and transport properties. The SVM classifier captures the heterogeneity in the whole core CT images more accurately compared with the core description, indicating that the CT images provide additional high-resolution information not observed by manual core description. Further, the obtained prediction results add information on the similarity of the lithofacies classes. The prediction results using the hybrid classifier are worse than the SVM classifier, indicating that low-power components may contain information that is required to differentiate among various lithofacies.

A novel approach to thermographic images analysis of equine thoracolumbar region: the effect of effort and rider\u2019s body weight on structural image complexity

BackgroundThe horses’ backs are particularly exposed to overload and injuries due to direct contact with the saddle and the influence of e.g. the rider’s body weight. The maximal load for a horse’s back during riding has been suggested not to exceed 20% of the horses’ body weight. The common prevalence of back problems in riding horses prompted the popularization of thermography of the thoracolumbar region. However, the analysis methods of thermographic images used so far do not distinguish loaded horses with body weight varying between 10 and 20%.ResultsThe superficial body temperature (SBT) of the thoracolumbar region of the horse’s back was imaged using a non-contact thermographic camera before and after riding under riders with LBW (low body weight, 10%) and HBW (high body weight, 15%). Images were analyzed using six methods: five recent SBT analyses and the novel approach based on Gray Level Co-Occurrence Matrix (GLCM) and Gray Level Run Length Matrix (GLRLM). Temperatures of the horse’s thoracolumbar region were higher (p < 0.0001) after then before the training, and did not differ depending on the rider’s body weight (p > 0.05), regardless of used SBT analysis method. Effort-dependent differences (p < 0.05) were noted for six features of GLCM and GLRLM analysis. The values of selected GLCM and GLRLM features also differed (p < 0.05) between the LBW and HBW groups.ConclusionThe GLCM and GLRLM analyses allowed the differentiation of horses subjected to a load of 10 and 15% of their body weights while horseback riding in contrast to the previously used SBT analysis methods. Both types of analyzing methods allow to differentiation thermal images obtained before and after riding. The textural analysis, including selected features of GLCM or GLRLM, seems to be promising tools in considering the quantitative assessment of thermographic images of horses’ thoracolumbar region.

Radiomic features of renal cell carcinoma primary and metastatic sites as predictors of TERT and BAP1 mutations.

282 Background: TERT and BAP1 mutations are associated with poor clinical outcome in patients (pts) with metastatic renal cell carcinoma (mRCC) (Dizman et al JITC 2020; Joseph et al J Urol 2016). In this study we explore radiogenomics as a non-invasive method to identify these alterations. Methods: Pts with mRCC who had genomic testing in the course of routine clinical care were included in the current analysis. Pts were assessed with the GEM Extra assay, a CAP-accredited, CLIA-certified test encompassing paired tumor-normal whole exome sequencing (WES) and tumor whole transcriptome sequencing (TGen; Phoenix, AZ). Pts underwent CT imaging; radiomic analysis was performed on the segmented metastatic and primary lesions. Features were independently correlated with TERT and BAP1 mutation status to generate Pearson correlation values (PCVs). Results: 92 pts (65:27 M: F) were included in the analysis; of these, the majority of pts (84%) had clear cell histology. Alterations in the TERT gene were seen in 12 pts. In these pts 1,325 radiomic features of the primary tissue were examined and 251 features correlated with a PCV ≥ |0.2|. Of these, 42 features were correlated with a PCV ≥ |0.3|. Highest correlation with TERT mutation was seen with Gray Level Cooccurrence Matrix (GLCM) and First Order Features (FOF). 9 pts had BAP1 mutation with 5 detected in primary tumor and 4 in metastatic sites. Analysis of primary tumor imaging yielded no significant associations between radiomic features and BAP1 mutation. However, out of approximately 1,500 radiomic features noted in metastatic sites, 111 features correlated with BAP1 mutation with a PCV ≥ 0.2. Of these, 15 features correlated with a PCV ≥ 0.3. The radiomic features with the highest correlation with BAP1mt were Gray Level Dependence Matrix (GLDM) and GLCM. Conclusions: By identifying a correlation between radiomic features of TERT mutation in primary tumors and BAP1 mutation in metastatic sites, our work may ultimately yield a non-invasive method of discerning mutational status in patents with mRCC. Efforts are ongoing to validate our findings within The Cancer Imaging Archive.

HPWO-LS-based deep learning approach with S-ROA-optimized optic cup segmentation for fundus image classification

Recently, automated retinal image processing has been considered a competitive field of research due to the low-accuracy results, complexity, and unacceptable outcomes associated with it. In this article, we proposed a novel approach for the classification of fundus images from different kinds of fundus disorders. The original images are preprocessed in terms of noise and contrast enhancement by using the contrast limited adaptive histogram equalization method. The optic cup segmentation from the fundus images is effectively handled via the search and rescue optimization algorithm. After that, the color, texture, and shape-based gray-level co-occurrence matrix features are extracted. The hybrid particle swarm optimization with local search strategy improves the DNN parameter and the newly developed method is named as optimal DNN. The optimal DNN is used to classify whether the image is diabetic retinopathy, glaucoma, or age-related macular degeneration. Experimentally, different kinds of datasets such as STARE, Drishti, and RIM-One datasets with performance measure are validated. Finally, the proposed approaches demonstrate higher classification performances in terms of accuracy, specificity, sensitivity, precision, recall, and f-measure.

Novel system in vitro of classifying oral carcinogenesis based on feature extraction for gray-level co-occurrence matrix using scanned laser pico projector.

Oral cancer is among the top 10 causes of death due to cancer worldwide. The prognosis for oral cancer patients is not good, with a 5-year survival rate of only 50%. Earlier and more precise classification will help clinicians make a diagnosis and patients survive. With the advancement of technology, computer-aided detection methods are used to help clinicians form therapy strategies. Gray-level co-occurrence matrix (GLCM) feature extraction of images describing the spatial distribution of gray levels is widely used in medical imaging analysis. Scanned laser pico projector (SLPP) has advantages such as high intensity, directivity, coherence, and mono-color with low bandwidth. In this study, GLCM feature extraction and SLPP reflex images were combined to make a small, non-staining, noninvasive classification system. According to the various image characteristics in oral carcinogenesis, SLPP reflex images better define the borders and three-dimensional structures and provide effective GLCM features such as contrast, energy, and homogeneity to classify carcinogenesis in dysplastic oral keratinocyte (DOK) and normal oral keratinocyte (NOK) cells. Moreover, it also reliably classifies highly metastatic (HSC-3) and tongue cancer (CAL-27) cells. A promising computer-aided classification system for oral cancer was developed to build a reliable intraoral examination system for in situ computer-aided diagnosis in normal clinics.

Unenhanced CT texture analysis with machine learning for differentiating between nasopharyngeal cancer and nasopharyngeal malignant lymphoma.

ABSTRACTDifferentiating between nasopharyngeal cancer and nasopharyngeal malignant lymphoma (ML) remains challenging on cross-sectional images. The aim of this study is to investigate the usefulness of texture features on unenhanced CT for differentiating between nasopharyngeal cancer and nasopharyngeal ML. Thirty patients with nasopharyngeal tumors, including 17 nasopharyngeal cancers and 13 nasopharyngeal MLs, were underwent 18F-FDG PET/CT. All nasopharyngeal cancers and 7 of 13 nasopharyngeal MLs were confirmed by endoscopic biopsy. On unenhanced CT, 34 texture features were analyzed following lesion segmentation in the maximum area of the target lesion. The Mann-Whitney U test and areas under the curve (AUCs) were used for analysis and to compare the maximum standardized uptake values (SUV)max, SUVmean, and 34 texture features. A support vector machine (SVM) was constructed to evaluate the diagnostic accuracy and AUCs of combinations of texture features, with 50 repetitions of 5-fold cross-validation. Differences between the SUVmax and SUVmean for nasopharyngeal cancers and nasopharyngeal MLs were not significant. Significant differences of texture features were seen, as follows: 1 histogram feature (p = 0.038), 3 gray-level co-occurrence matrix features (p < 0.05), and 1 neighborhood gray-level different matrix feature (NGLDM) (p = 0.003). Coarseness in NGLDM provided the highest diagnostic accuracy and largest AUC of 76.7% and 0.82, respectively. SVM evaluation of the combined texture features obtained the highest accuracy of 81.3%, with an AUC of 0.80. Combined texture features can provide useful information for discriminating between nasopharyngeal cancer and nasopharyngeal ML on unenhanced CT.

An efficient wavelet-based image fusion for brain tumor detection and segmentation over PET and MRI image

A brain tumor is an abnormal growth of cells, reproducing themselves in an uncontrolled manner. In a medical diagnosis system, the accurate detection of location and size plays a very important role in the diagnosis of brain tumors. Magnetic Resonance Imaging (MRI) and Positron Emission Tomography (PET) are the most widely used techniques for diagnosis. The detection of a brain tumor in PET and MRI images is a challenging task due to its low sensitive boundary pixels. For accurate detection of brain tumors, in this paper, an efficient fusion-based brain tumor detection and segmentation is proposed. Here, at first, we fuse the input image using a discrete wavelet transform (DWT) and novel fusion rule. After the fusion process, the gray-level co-occurrence matrix (GLCM) features are extracted. Then, we categorize the brain images as normal and abnormal images using Optimal Deep Neural Network (ODNN). Here, the network weights of DNN are optimally selected using Spider Monkey Optimization (SMO) algorithm. After the classification process, the brain tumor region is extracted from abnormal brain images using the weighted k-means technique. The performance of the proposed methodology is analyzed in terms of sensitivity, specificity, and accuracy.

Gray Level Co-Occurrence Matrix (GLCM) and Gabor Features Based No-Reference Image Quality Assessment for Wood Images

Gray Level Co-Occurrence Matrix (GLCM) and Gabor Features Based No-Reference Image Quality Assessment for Wood Images

Wayang Image Classification Using SVM Method and GLCM Feature Extraction

Wayang is a masterpiece of art that has been able to survive centuries of change and development as a reflection of life for the majority of society. Wayang has a high value because it does not only function as a "entertainment" spectacle, but also has many lessons and life values that can be learned from a wayang show. Puppet itself has various types and forms, and these forms have their own uniqueness, because of the many types of Puppet, many people do not know all the names and types of wayang. Therefore, in this research, we will discuss how to recognize wayang objects based on wayang images using the SVM and GLCM methods as feature extraction. The results showed that the classification of wayang using the SVM (Support Vector Machine) method and the GLCM (Gray Level Co-Occurrence Matrix) feature extraction can recognize wayang objects based on wayang images and classify them quite accurately and a maximum total accuracy of 83.2% is obtained.

Reliability of radiomics features due to image reconstruction using a standardized T2 -weighted pulse sequence for MR-guided radiotherapy: An anthropomorphic phantom study.

To prospectively investigate the impact of image reconstruction on MRI radiomics features. An anthropomorphic phantom was scanned at 1.5 T using a standardized sequence for MR-guided radiotherapy under SENSE and compressed-SENSE reconstruction settings. A total of 93 first-order and texture radiomics features in 10 volumes of interest were assessed based on (1) accuracy measured by the percentage deviation from the reference, (2) robustness on reconstruction in all volumes of interest measured by the intraclass correlation coefficient, and (3) repeatability measured by the coefficient of variance over the repetitive acquisitions. Finally, reliable and unreliable radiomics features were comprehensively determined based on their accuracy, robustness, and repeatability. Better accuracy and robustness of the radiomics features were achieved under SENSE than compressed-SENSE reconstruction. The feature accuracy under SENSE reconstruction was more affected by acceleration factor than direction, whereas under compressed-SENSE reconstruction, accuracy was substantially impacted by the increasing denoising levels. Feature repeatability was dependent more on feature types than on reconstruction. A total of 45 reliable features and 13 unreliable features were finally determined for SENSE, compared with 22 reliable and 26 unreliable features for compressed SENSE. First-order and gray-level co-occurrence matrix features were generally more reliable than other features. Radiomics features could be substantially affected by MRI reconstruction, so precautions need to be taken regarding their reliability for clinical use. This study helps the guidance of the preselection of reliable radiomics features and the preclusion of unreliable features in MR-guided radiotherapy.

Improved Rough-fuzzy C-means Clustering and Optimum Fuzzy Interference System for MRI Brain Image Segmentation

The categorization of brain tissues plays a vital role in various neuro-anatomical identification and implementations. In manual detection, misidentification of location and sound of unwanted tissues may occur due to visual fatigue by humans. Also, it consumes more time and may exhibit enormous partially inner or outer the manipulator. At present, automatic identification of brain tissues in MRI is vital for investigation and healing applications. This work proposed MRI image tissue segmentation using Improved Rough Fuzzy C Means (IRFCM) algorithm and classification using multiple fuzzy systems. Proposed research work comprises four modules: pre-processing, segmentation, categorization, and extracting features. Initially, the elimination of boisterous occur in the given image is done through pre-processing. After the pre-processing, segmentation is carried out for the pre-processed brain image to segment the tissue based on clustering concept using Improved Rough Fuzzy C Means algorithm. Later, the features of Gray-Level Co-Occurrence Matrix (GLCM) are extracted from segmentation, and the features extracted from segmented images are applied to Optimum Fuzzy Interference System (OFIS). Then the entire system parameters are optimized using Enhanced Grasshopper Optimization Algorithm (EGOA). Finally, the novel OFIS classifier helps to classify the brain-based tissue images as Gray Matter (GM), White Matter (WM), Cerebrospinal Fluid (CSF), and Tumor Tissues (TT). The results using MRI data sets are analyzed and compared with other existing techniques through performance metrics to show the superiority of the proposed methodology.

Offline handwritten signature verification using various Machine Learning Algorithms

In today’s world it is necessary to protect one’s authenticity in order to ensure the protection of personal information that only the authenticate credentials of a person can have access to. Nowadays there is an increase in number of malpractices like signature forgery to access the important information of a person. To encounter signature verification problem, there have been a number of advances in verifying the authenticity of signature using various techniques including Machine Learning and Deep Learning. This paper introduces a novel approach to verify the signatures using difference of gaussian filtering technique, gray level co-occurrence matrix feature extraction technique, principle component analysis and kernel principal component analysis associated with various machine learning algorithms. The publicly available Kaggle offline handwritten signature dataset is used for training. This article compares the accuracy of the dataset on various machine learning algorithms. After training datasets the lowest accuracy achieved is 56.66% for Naive Bayes algorithm. The highest accuracy achieved is 82% for K-Nearest Neighbour (KNN) and 81.66% for Random Forest using principle components and kernel principle components of the dataset.

Time-Frequency Analysis of EEG Signals and GLCM Features for Depth of Anesthesia Monitoring.

One of the important tasks in the operating room is monitoring the depth of anesthesia (DoA) during surgery, and noninvasive techniques are very popular. Hence, we propose a new scheme for DoA monitoring considering the time-frequency analysis of electroencephalography (EEG) signals and GLCM features extracted from them. To this end, at first, the time-frequency map (TFM) of each channel of each EEG is computed by smoothed pseudo-Wigner–Ville distribution (SPWVD), where the EEG signal used in this paper is recorded in 15 channels. After that, we consider the gray-level co-occurrence matrix (GLCM) to obtain the content of TFM, and after that, four features such as homogeneity, correlation, energy, and contrast are obtained for each GLCM. Finally, after the selection of efficient features using the minimum redundancy maximum relevance (MRMR) method, the K-nearest neighbor (KNN) classifier is utilized to determine the DoA. Here, we consider the three states, namely, deep hypnotic, surgical anesthesia, and sedation and awake states according to bispectral index (BIS), and each EEG epoch is classified to these states. We also employ data augmentation to enhance the training phase and increase accuracy. We obtain the accuracy and confusion matrix of the proposed method. We also analyze the effects of a number of gray levels of GLCM, distance measure in KNN classifier, and parameters of data augmentation on the performance of the proposed method. Results indicate the efficiency of the proposed method to determine the DoA during surgery.

Digital Forensics for Skulls Classification in Physical Anthropology Collection Management

The size, shape, and physical characteristics of the human skull are distinct when considering individual humans. In physical anthropology, the accurate management of skull collections is crucial for storing and maintaining collections in a cost-effective manner. For example, labeling skulls inaccurately or attaching printed labels to skulls can affect the authenticity of collections. Given the multiple issues associated with the manual identification of skulls, we propose an automatic human skull classification approach that uses a support vector machine and different feature extraction methods such as gray-level co-occurrence matrix features, Gabor features, fractal features, discrete wavelet transforms, and combinations of features. Each underlying facial bone exhibits unique characteristics essential to the face's physical structure that could be exploited for identification. Therefore, we developed an automatic recognition method to classify human skulls for consistent identification compared with traditional classification approaches. Using our proposed approach, we were able to achieve an accuracy of 92.3–99.5% in the classification of human skulls with mandibles and an accuracy of 91.4–99.9% in the classification of human skills without mandibles. Our study represents a step forward in the construction of an effective automatic human skull identification system with a classification process that achieves satisfactory performance for a limited dataset of skull images.

A Two-Scale Method of Sea Ice Classification Using TerraSAR-X ScanSAR Data During Early Freeze-Up

Sea ice classification using synthetic aperture radar (SAR) in the wide-swath mode is one method for ice type retrieval. The conventional method of sea ice classification involves single-scale classification, wherein the backscatter and gray-level co-occurrence matrix features in a specific resolution are used as classification bases. The classification is implemented only once. In this way, however, the scalloping and interscan banding (ISB) artifacts in some TerraSAR-X ScanSAR data affect the classification result. In this article, a two-scale method of sea ice classification is proposed. The two-scale process includes both the coarse- and fine-resolution classifications. First, the coarse-resolution classification utilizes downscale images to generate the ice zone. Subsequently, the fine-scale classification utilizes the full-resolution SAR image and the ice zone to generate the final classification result. The support vector machine is used as the classifier. This two-scale method provides a better classification result compared to the single-scale method in two aspects. First, it effectively avoids the appearance of scalloping and ISB artifacts in the results. Second, it is less time-consuming and provides satisfactory results for discriminating multiyear ice, first-year ice, and water. The validation is accomplished using independent images, with an overall average accuracy of 86.38%. This article shows that this method may provide an alternative for sea ice classification using TerraSAR-X ScanSAR data.

Artificial intelligence-based myocardial texture analysis in etiological differentiation of left ventricular hypertrophy.

BackgroundTransthoracic echocardiography (TTE) is widely used in clinics to evaluate left ventricular hypertrophy (LVH). However, TTE is usually insufficient for the etiological diagnoses when morphological and functional features are nonspecific. With the booming of computer science and artificial intelligence (AI), previous literature has reported the application of radiomics based on cardiac magnetic resonance imaging, cardiac computed tomography and TTE in diagnosing several myocardial abnormalities, such as myocardial infarction, myocarditis, cardiac amyloidosis, and hypertrophic cardiomyopathy (HCM). In this study, we explored the possibility of using myocardial texture features in differentiating HCM, hypertensive heart disease (HHD) and uremic cardiomyopathy (UCM) based on echocardiography. To our knowledge, this was the first study to explore TTE myocardial texture analysis for multiple LVH etiology differentiation.MethodsTTE images were reviewed retrospectively from January 2018 to collect 50 cases for each group of HHD, HCM and UCM. The apical four chamber view was retrieved. Seventeen first-order statistics and 60 gray level co-occurrence matrix (GLCM) features were extracted for statistics and classification test by support vector machine (SVM).ResultsOf all the parameters, entropy of brightness (EtBrt), standard deviation (Std), coefficient of variation (CoV), skewness (Skew), contrast7 (Cont7) and homogeneity5 (Hm5) were found statistically significant among the three groups (all P<0.05) and with acceptable reproducibility (intraobserver and interobserver ICC >0.50). As a result, HCM showed the most homogeneous myocardial texture, and was significantly different from HHD and UCM (all six features: P≤0.005). HHD appeared slightly more homogeneous than UCM, as only EtBrt and CoV were significant (P=0.011 and P=0.008). According to higher areas under the receiver operating characteristic curve (AUC) (>0.50), EtBrt, Std, and CoV were selected for test of classification as a combination of features. The AUC derived from SVM model was slightly improved compared with those of EtBrt, Std and CoV individually.ConclusionsAI-based myocardial texture analysis using ultrasonic images may be a potential approach to aiding LVH etiology differentiation.

Diagnosis of COVID-19 Infection Using Three-Dimensional Semantic Segmentation and Classification of Computed Tomography Images

Coronavirus 19 (COVID-19) can cause severe pneumonia that may be fatal. Correct diagnosis is essential. Computed tomography (CT) usefully detects symptoms of COVID-19 infection. In this retrospective study, we present an improved framework for detection of COVID-19 infection on CT images; the steps include pre-processing, segmentation, feature extraction/fusion/selection, and classification. In the pre-processing phase, a Gabor wavelet filter is applied to enhance image intensities. A marker-based, watershed controlled approach with thresholding is used to isolate the lung region. In the segmentation phase, COVID-19 lesions are segmented using an encoder-/decoder-based deep learning model in which deepLabv3 serves as the bottleneck and mobilenetv2 as the classification head. DeepLabv3 is an effective decoder that helps to refine segmentation of lesion boundaries. The model was trained using fine-tuned hyperparameters selected after extensive experimentation. Subsequently, the Gray Level Co-occurrence Matrix (GLCM) features and statistical features including circularity, area, and perimeters were computed for each segmented image. The computed features were serially fused and the best features (those that were optimally discriminatory) selected using a Genetic Algorithm (GA) for classification. The performance of the method was evaluated using two benchmark datasets: The COVID-19 Segmentation and the POF Hospital datasets. The results were better than those of existing methods.

Ensemble learning-based classification on local patches from magnetic resonance images to detect iron depositions in the brain

Iron deposition in the brain has been observed with normal aging and is associated with neurodegenerative diseases. The automated classification of brain magnetic resonance images (MRI) based on iron deposition in basal ganglia region of the brain has not been performed, to our knowledge. It is very difficult to analyse iron regions in brain using simple MRI techniques. The MRI sequence namely susceptibility weighted imaging (SWI) helps to distinguish brain iron regions. The objective of our work is to investigate the iron regions in selected areas of basal ganglia region of brain and classify MR images. The study included a total of 60 MRI images which consists of 40 subjects with iron region and 20 subjects of healthy controls. We performed Gaussian smoothing followed by construction of 40 localised patches of each MR image based on iron and normal regions. Grey level co-occurrence matrix (GLCM) features are extracted from the patches and fed to random forest (RF) classifier for patch-based classification of iron region. Training of data patch features was done by random forest classifier and the performance of classifier in terms of accuracy was measured. The experimental results show that the proposed localised patch-based approach for classification of brain iron images using random forest classifier achieved 96.25% classification accuracy in identifying normal and iron regions from brain MR sequences.

Ensemble learning-based classification on local patches from magnetic resonance images to detect iron depositions in the brain

Iron deposition in the brain has been observed with normal aging and is associated with neurodegenerative diseases. The automated classification of brain magnetic resonance images (MRI) based on iron deposition in basal ganglia region of the brain has not been performed, to our knowledge. It is very difficult to analyse iron regions in brain using simple MRI techniques. The MRI sequence namely susceptibility weighted imaging (SWI) helps to distinguish brain iron regions. The objective of our work is to investigate the iron regions in selected areas of basal ganglia region of brain and classify MR images. The study included a total of 60 MRI images which consists of 40 subjects with iron region and 20 subjects of healthy controls. We performed Gaussian smoothing followed by construction of 40 localised patches of each MR image based on iron and normal regions. Grey level co-occurrence matrix (GLCM) features are extracted from the patches and fed to random forest (RF) classifier for patch-based classification of iron region. Training of data patch features was done by random forest classifier and the performance of classifier in terms of accuracy was measured. The experimental results show that the proposed localised patch-based approach for classification of brain iron images using random forest classifier achieved 96.25% classification accuracy in identifying normal and iron regions from brain MR sequences.