Gray Level Co-occurrence Matrix Features Research Articles

Identification of healthy biological leafs using hybrid-feature classifier

ABSTRACT Plants are an important element of the ecosystem that helps in controlling carbon emissions and environmental changes. Characterization and identification are a need for protecting plants and for people to understand plant protection. Plant leaves are the main parts for detection. Characterizing leaves now has been a significant and complicated task, particularly with the features of leaves. Leaf images of two different types are considered here, one is healthy while the other is unhealthy, and divided into two distinct classes. The proposed method incorporates features of the leaf images that are extracted utilizing the Gray Level Co-occurrence Matrix (GLCM) and Gray Level Run Length Matrix (GLRLM) feature extraction techniques. The outcomes are classified using three different classifiers: Random Forest, Multilayer Perceptron, and Naïve Bayes with an accuracy of 95.84%, 98.33%, and 82.89% respectively. The classifiers successfully classify the healthy and diseased leaves of various plants that were considered here. Hence as per the investigation, the study can be valuable for analysts for plant recognition, characterization, and diagnosis.

Radiomics signature for temporal evolution and recurrence patterns of glioblastoma using multimodal magnetic resonance imaging.

Glioblastoma is a highly infiltrative neoplasm with a high propensity of recurrence. The location of recurrence usually cannot be anticipated and depends on various factors, including the surgical resection margins. Currently, radiation planning utilizes the hyperintense signal from T2-FLAIR MRI and is delivered to a limited area defined by standardized guidelines. To this end, noninvasive early prediction and delineation of recurrence can aid in tailored targeted therapy, which may potentially delay the relapse, consequently improving overall survival. In this work, we hypothesize that radiomics-based phenotypic quantifiers may support the detection of recurrence before it is visualized on multimodal MRI. We employ retrospective longitudinal data from 29 subjects with a varying number of time points (three to 13) that includes glioblastoma recurrence. Voxelwise textural and intensity features are computed from multimodal MRI (T1-contrast enhanced [T1CE], FLAIR, and apparent diffusion coefficient), primarily to gain insights into longitudinal radiomic changes from preoperative MRI to recurrence and subsequently to predict the region of relapse from 143 ± 42 days before recurrence using machine learning. T1CE MRI first-order and gray-level co-occurrence matrix features are crucial in detecting local recurrence, while multimodal gray-level difference matrix and first-order features are highly predictive of the distant relapse, with a voxelwise test accuracy of 80.1% for distant recurrence and 71.4% for local recurrence. In summary, our work exemplifies a step forward in predicting glioblastoma recurrence using radiomics-based phenotypic changes that may potentially serve as MR-based biomarkers for customized therapeutic intervention.

Flood damage assessment with Sentinel-1 and Sentinel-2 data after Sardoba dam break with GLCM features and Random Forest method

Accurate mapping and monitoring of flooded areas are immensely required for disaster management purposes, such as for damage assessment and mitigation. In this study, the flood damage mapping performances of two satellite Earth Observation sensors, i.e., European Space Agency's Sentinel-1 (S1) synthetic aperture radar (SAR) and Sentinel-2 (S2) multispectral optical instruments, were evaluated using the Random Forest (RF) supervised classification method and various feature types. The study area was Sardoba Reservoir (Uzbekistan) and its surroundings, in which a disastrous dam failure occurred on May 1, 2020. After the failure of a part of the earthfill dam, a large region with settlements and agricultural areas in Uzbekistan and Kazakhstan was flooded. S1 and S2 cloudless data with a short temporal interval acquired soon after the event were available for the area. Four different data availability scenarios, such as (i) only S1 pre- and post-flood data; (ii) only S2 pre- and post-flood data; (iii) S1 pre- and post-flood and S2 pre-flood data; and (iv) S1 and S2 pre- and post-flood data were evaluated in terms of classification accuracy. In addition to the polarization information of S1 and the intensity values of S2 bands, feature maps produced from these datasets, such as vegetation and water indices, textural information obtained from gray level co-occurrence matrix (GLCM), and the principal component analysis (PCA) bands were employed in the RF method. The results show that the fusion of S1 and S2 data exhibit very high classification accuracy for the flooded areas and can separate the inundated vegetation as well. The use of S2 pre-event data together with the S1 pre- and post-event data is recommended for obtaining high accuracy even when post-event optical data is not available.

Detection of Iris Presentation Attacks Using Feature Fusion of Thepade's Sorted Block Truncation Coding with Gray-Level Co-Occurrence Matrix Features.

Iris biometric detection provides contactless authentication, preventing the spread of COVID-19-like contagious diseases. However, these systems are prone to spoofing attacks attempted with the help of contact lenses, replayed video, and print attacks, making them vulnerable and unsafe. This paper proposes the iris liveness detection (ILD) method to mitigate spoofing attacks, taking global-level features of Thepade’s sorted block truncation coding (TSBTC) and local-level features of the gray-level co-occurrence matrix (GLCM) of the iris image. Thepade’s SBTC extracts global color texture content as features, and GLCM extracts local fine-texture details. The fusion of global and local content presentation may help distinguish between live and non-live iris samples. The fusion of Thepade’s SBTC with GLCM features is considered in experimental validations of the proposed method. The features are used to train nine assorted machine learning classifiers, including naïve Bayes (NB), decision tree (J48), support vector machine (SVM), random forest (RF), multilayer perceptron (MLP), and ensembles (SVM + RF + NB, SVM + RF + RT, RF + SVM + MLP, J48 + RF + MLP) for ILD. Accuracy, precision, recall, and F-measure are used to evaluate the performance of the projected ILD variants. The experimentation was carried out on four standard benchmark datasets, and our proposed model showed improved results with the feature fusion approach. The proposed fusion approach gave 99.68% accuracy using the RF + J48 + MLP ensemble of classifiers, immediately followed by the RF algorithm, which gave 95.57%. The better capability of iris liveness detection will improve human–computer interaction and security in the cyber-physical space by improving person validation.

Role of 18F-FDG PET/CT Radiomics Features in the Differential Diagnosis of Solitary Pulmonary Nodules: Diagnostic Accuracy and Comparison between Two Different PET/CT Scanners.

The aim of this retrospective study was to investigate the ability of 18 fluorine-fluorodeoxyglucose positron emission tomography/CT (18F-FDG-PET/CT) metrics and radiomics features (RFs) in predicting the final diagnosis of solitary pulmonary nodules (SPN). We retrospectively recruited 202 patients who underwent a 18F-FDG-PET/CT before any treatment in two PET scanners. After volumetric segmentation of each lung nodule, 8 PET metrics and 42 RFs were extracted. All the features were tested for significant differences between the two PET scanners. The performances of all features in predicting the nature of SPN were analyzed by testing three classes of final logistic regression predictive models: two were built/trained through exploiting the separate data from the two scanners, and the other joined the data together. One hundred and twenty-seven patients had a final diagnosis of malignancy, while 64 were of a benign nature. Comparing the two PET scanners, we found that all metabolic features and most of RFs were significantly different, despite the cross correlation being quite similar. For scanner 1, a combination between grey level co-occurrence matrix (GLCM), histogram, and grey-level zone length matrix (GLZLM) related features presented the best performances to predict the diagnosis; for scanner 2, it was GLCM and histogram-related features and metabolic tumour volume (MTV); and for scanner 1 + 2, it was histogram features, standardized uptake value (SUV) metrics, and MTV. RFs had a significant role in predicting the diagnosis of SPN, but their accuracies were directly related to the scanner.

Predictive Ability of Delta Radiomic Texture Features (DRTF) Extracted from Liver Patients Treated With Magnetic Resonance Guided Stereotactic Body Radiotherapy (MRgSBRT)

In the abdomen, real-time MRgSBRT has live motion tracking, high soft tissue contrast, and direct visualization of oncologic targets and organs at risk. Setup images are acquired daily, which generates longitudinal, non-invasive radiographic data points. Our study aimed to identify delta radiomic texture features (DRTF) extracted from daily set up images to predict local progression in liver patients treated with MRgSBRT.This work included a prospective analysis of an IRB-approved database of 20 patients with liver lesions treated by MRgSBRT. The gross tumor volume was identified on daily low field strength (0.35T) setup images. MR images were pre-processed using histogram equalization to re-bin the gray level intensities to 64 levels and 39 second-order DRTFs extracted. DRTFs were calculated as the difference between values on the initial set up scan and after intervals of BED 20 Gy (α/β = 10) were delivered. The Gini Index was used to rank features in order of predictive importance during training of a random forest model while using the delta radiomics texture features to predict local control (LC) at 1 year. Leave-one-out (LOO) analysis was performed using the top two features.Lesions were treated to a median dose of 50 Gy in 5 fractions, or BED 20 Gy per fraction. The top two features identified after BED 20 Gy (1 fraction) was delivered were gray level co-occurrence matrix features (GLCM) and gray level size zone matrix (GLSZM) based Large Zone Emphasis. The model generated an ROC curve with AUC = 0.9011 [0.752 - 1] using LOO analysis. The same two features were selected as predictive after delivery of BED 40 Gy and LOO analysis resulted in a ROC curve with AUC = 0.6868 [0.3957 - 0.9779].DRTFs extracted from daily setup MR images after a single fraction of MRgSBRT were predictive of LC in this cohort. Additional studies are needed to evaluate these features as they may identify patients with radioresistant disease and provide an opportunity for physicians to dose escalate.

A Robust Machine Learning Approach for Multiclass Alzheimer’s Disease Detection using 3D Brain Magnetic Resonance Images

Alzheimer’s disease (AD), a progressive dementia is the neurodegenerative disorder that worsens memory and mental capabilities mostly in aged people. Currently, clinical and psychometric assessments are being used to diagnose the disease in patients. In clinical procedures, 3D Magnetic Resonance Image qualitative parameters are analyzed to identify the abnormality in brain shape, volume, texture, and cortical thickness. This paper presents a robust approach for categorizing 3D MR images into multiple stages of AD using hybrid features viz., Gray Level Co-occurrence Matrix (GLCM), 3D Scale and rotation Invariant Feature Transform (3D SIFT), Histogram of Oriented Gradients – Three Orthogonal Planes (HOG-TOP) and Complete Local Binary Pattern of Sign and Magnitude – Three Orthogonal Planes (CLBPSM-TOP). The proposed algorithm is validated using Open Access Series of Imaging Studies (OASIS) datasets to classify the subjects into AD, Mild Cognitive Impairment (MCI) and Cognitive Normal (CN) categories using various classifiers. Moreover, this approach is also evaluated and compared with the state-of-the-art approaches. 87.84% diagnosis accuracy is achieved with Ensemble classifier using hybrid features to diagnose the severity of AD. This approach also outperforms majority of these techniques in key parameters viz., accuracy, precision, recall and F1-score.

Magnetic resonance imaging texture analysis in the quantitative evaluation of acute osteomyelitis of the mandibular bone.

Accurate assessment of radiological images can help in early diagnosis and therapy of suppurative osteomyelitis (OM). The purpose of this study was to apply texture analysis to MRI as a means of quantitatively evaluating acute OM of the mandible. We analyzed the data from 38 patients who complained of pain and underwent MRI between April 2017 and March 2019. From the MRIs of these patients, with (n = 19) and without OM (n = 19), 279 radiomics features were extracted using short tau inversion recovery, data of the regions of interest and analyzed with MaZda v. 3.3. 10 features, including one histogram feature (90th percentile), eight gray-level co-occurrence matrix features (Sum Averg), and one gray-level run-length matrix feature (Horzl_RLNonUni), were selected using Fisher coefficient and compared between the acute OM and non-OM groups. The two groups were compared using Mann-Whitney U test with p value set at 0.05. All 10 radiomics features showed significant differences between the acute OM and non-OM groups (p < 0.05). MRI texture analysis has potential application in radiomics diagnosis of acute OM of the mandible.

Lung Nodule Detection of CT and Image-Based GLCM and RLM CT Scan Using the Support Vector Machine (SVM) Method

Lung cancer is the most common cause of cancer death globally. Early detection of lung cancer will greatly beneficial to save the patient. This study focused on the detection of lung cancer using classification with the Support Vector Machine (SVM) method based on the features of Gray Level Co-occurrence Matrices (GLCM) and Run Length Matrix (RLM). The lung data used were obtained from the Cancer imaging archive Database, consisting of 500 CT images. CT images were grouped into 2 clusters, including normal and lung cancer. The research steps include: image processing, region of interest segmentation, and feature extraction. The results indicate that the system can detect the CT-image of SVM classification where the default parameter only provides an accuracy of 85.63%. It is expected that the results will be useful to help medical personnel and researchers to detect the status of lung cancer. These results provide information that detection of lung nodules based on GLCM and RLM features that can be detected is better. Furthermore, selecting parameters C and γ on SVM. Keywords: cancer, nodule, support vector machine (SVM).

On dose cube pixel spacing pre-processing for features extraction stability in dosiomic studies.

Dosiomics allows to parameterize regions of interest (ROIs) and to produce quantitative dose features encoding the spatial and statistical distribution of radiotherapy dose. The stability of dosiomics features extraction on dose cube pixel spacing variation has been investigated in this study. Based on 17 clinical delivered dose distributions (Pn), dataset has been generated considering all the possible combinations of four dose grid resolutions and two calculation algorithms. Each dose voxel cube has been post-processed considering 4 different dose cube pixel spacing values: 1x1x1, 2x2x2, 3x3x3 mm3 and the one equal to the planning CT. Dosiomics features extraction has been performed from four different ROIs. The stability of each extracted dosiomic feature has been analyzed in terms of coefficient of variation (CV) intraclass correlation coefficient (ICC). The highest CV mean values were observed for PTV ROI and for the grey level size zone matrix features family. On the other hand, the lowest CV mean values have been found for RING ROI for the grey level co-occurrence matrix features family. P3 showed the highest percentage of CV>1 (1.14%) followed by P15 (0.41%), P1 (0.29%) and P13 (0.19%). ICC analysis leads to identify features with an ICC>0.95 that could be considered stable to use in dosiomic studies when different dose cube pixel spacing are considered, especially the features in common among the seventeen plans. Considering the observed variability, dosiomic studies should always provide a report not only on grid resolution and algorithm dose calculation, but also on dose cube pixel spacing.

Thermal Imagery Feature Extraction Techniques and the Effects on Machine Learning Models for Smart HVAC Efficiency in Building Energy

The control of thermostats of a heating, ventilation, and air-conditioning (HVAC) system installed in commercial and residential buildings remains a pertinent problem in building energy efficiency and thermal comfort research. The ability to determine the number of people at a particular time in an area is imperative for energy efficiency in order to condition only occupied regions and thermally deficient regions. In this study of the best features comparison for detecting the number of people in an area, feature extraction techniques including wavelet scattering, wavelet decomposition, grey-level co-occurrence matrix (GLCM) and feature maps convolution neural network (CNN) layers were explored using thermal camera imagery. Specifically, the pretrained CNN networks explored are the deep residual (Resnet-50) and visual geometry group (VGG-16) networks. The discriminating potential of Haar, Daubechies and Symlets wavelet statistics on different distributions of data were investigated. The performance of VGG-16 and ResNet-50 in an end-to-end manner utilizing transfer learning approach was investigated. Experimental results showed the classification and regression trees (CART) model trained on only GLCM and Haar wavelet statistic features, individually achieved accuracies of approximately 80% and 84%, respectively, in the detection problem. Moreover, k-nearest neighbors (KNN) trained on the combined features of GLCM and Haar wavelet statistics achieved an accuracy of approximately 86%. In addition, the performance accuracy of the multi classification support vector machine (SVM) trained on deep features obtained from layers of pretrained ResNet-50 and VGG-16 was between 96% and 97%. Furthermore, ResNet-50 transfer learning outperformed the VGG-16 transfer learning model for occupancy detection using thermal imagery. Overall, the SVM model trained on features extracted from wavelet scattering emerged as the best performing classifier with an accuracy of 100%. A principal component analysis (PCA) on the wavelet scattering features proved that the first twenty (20) principal components achieved a similar accuracy level instead of training on the whole feature set to reduce the execution time. The occupancy detection models can be integrated into HVAC control systems for energy efficiency and security systems, and aid in the distribution of resources to people in an area.

The Detection of Brain Tumors Using Chan-Vese Active Contour Without Edges Method in Magnetic Resonance (MR) Images

Accurate and automatic detections of brain tumors are vital. The aim of this study is to detect brain tumors in Magnetic Resonance (MR) images and to classify these tumors with a high degree of accuracy. After removing skull, the suspicious regions including tumors in the MR images were detected by using K-means clustering, K-means clustering in Lab color space, and the Chan-Vese without edges algorithm. At this stage, a performance evaluation of these three different methods was investigated, and it was seen that the best result was obtained in the Chan-Vese active contour without edges algorithm. For the classification stage, various features such as shape-based features, gray level co-occurrence matrix features, histogram of oriented gradients features, local binary pattern features, and statistical features were extracted from the detected suspicious regions. Finally, the suspicious regions were classified by k-nearest neighbor (k-NN), Fisher’s linear discriminant analysis (FLDA), random forest, decision tree, support vector machines (SVM), logistic linear classifier (LLC), and Naive Bayes classification methods. As a result of this study, it was determined that the FLDA classifier provided the best results with 93.01% accuracy, 93.46% sensitivity, and 96.50% specificity rates in classification for benign tumors, malignant tumors, and healthy (without tumor) cases.

Radiomic Values from High-Grade Subtypes to Predict Spread Through Air Spaces in Lung Adenocarcinoma

BackgroundWe aimed to establish a radiomic prediction model for tumor spread through air spaces (STAS) in lung adenocarcinoma using radiomic values from high-grade subtypes (solid and micropapillary). MethodsWe retrospectively reviewed 327 patients with lung adenocarcinoma from 2 institutions (cohort 1: 227 patients; cohort 2: 100 patients) between March 2017 and March 2019. STAS was identified in 113 (34.6%) patients. A high-grade likelihood prediction model was constructed based on a historical cohort of 82 patients with “near-pure” pathologic subtype. The STAS prediction model based on the patch-wise mechanism identified the high-grade likelihood area for each voxel within the internal border of the tumor. STAS presence was indirectly predicted by a volume percentage threshold of the high-grade likelihood area. Performance was evaluated by receiver operating curve analysis with 10-repetition, 3-fold cross-validation in cohort 1, and was individually tested in cohort 2. ResultsOverall, 227 patients (STAS-positive: 77 [33.9%]) were enrolled for cross-validation (cohort 1) while 100 (STAS-positive: 36 [36.0%]) underwent individual testing (cohort 2). The gray level cooccurrence matrix (variance) and histogram (75th percentile) features were selected to construct the high-grade likelihood prediction model, which was used as the STAS prediction model. The proposed model achieved good performance in cohort 1 with an area under the curve, sensitivity, and specificity, of 81.44%, 86.75%, and 62.60%, respectively, and correspondingly, in cohort 2, they were 83.16%, 83.33%, and 63.90%, respectively. ConclusionsThe proposed computed tomography–based radiomic prediction model could help guide preoperative prediction of STAS in early-stage lung adenocarcinoma and relevant surgeries.

Apple quality identification and classification by image processing based on convolutional neural networks

This work researched apple quality identification and classification from real images containing complicated disturbance information (background was similar to the surface of the apples). This paper proposed a novel model based on convolutional neural networks (CNN) which aimed at accurate and fast grading of apple quality. Specific, complex, and useful image characteristics for detection and classification were captured by the proposed model. Compared with existing methods, the proposed model could better learn high-order features of two adjacent layers that were not in the same channel but were very related. The proposed model was trained and validated, with best training and validation accuracy of 99% and 98.98% at 2590th and 3000th step, respectively. The overall accuracy of the proposed model tested using an independent 300 apple dataset was 95.33%. The results showed that the training accuracy, overall test accuracy and training time of the proposed model were better than Google Inception v3 model and traditional imaging process method based on histogram of oriented gradient (HOG), gray level co-occurrence matrix (GLCM) features merging and support vector machine (SVM) classifier. The proposed model has great potential in Apple’s quality detection and classification.

Radiomic Score as a Potential Imaging Biomarker for Predicting Survival in Patients With Cervical Cancer.

ObjectivesAccurate prediction of prognosis will help adjust or optimize the treatment of cervical cancer and benefit the patients. We aimed to investigate the incremental value of radiomics when added to the FIGO stage in predicting overall survival (OS) in patients with cervical cancer.MethodsThis retrospective study included 106 patients with cervical cancer (FIGO stage IB1–IVa) between October 2017 and May 2019. Patients were randomly divided into a training cohort (n = 74) and validation cohort (n = 32). All patients underwent contrast-enhanced computed tomography (CT) prior to treatment. The ITK-SNAP software was used to delineate the region of interest on pre-treatment standard-of-care CT scans. We extracted 792 two-dimensional radiomic features by the Analysis Kit (AK) software. Pearson correlation coefficient analysis and Relief were used to detect the most discriminatory features. The radiomic signature (i.e., Radscore) was constructed via Adaboost with Leave-one-out cross-validation. Prognostic models were built by Cox regression model using Akaike information criterion (AIC) as the stopping rule. A nomogram was established to individually predict the OS of patients. Patients were then stratified into high- and low-risk groups according to the Youden index. Kaplan–Meier curves were used to compare the survival difference between the high- and low-risk groups.ResultsSix textural features were identified, including one gray-level co-occurrence matrix feature and five gray-level run-length matrix features. Only the FIGO stage and Radscore were independent risk factors associated with OS (p < 0.05). The C-index of the FIGO stage in the training and validation cohorts was 0.703 (95% CI: 0.572–0.834) and 0.700 (95% CI: 0.526–0.874), respectively. Correspondingly, the C-index of Radscore was 0.794 (95% CI: 0.707–0.880) and 0.754 (95% CI: 0.623–0.885). The incorporation of the FIGO stage and Radscore achieved better performance, with a C-index of 0.830 (95% CI: 0.738–0.922) and 0.772 (95% CI: 0.615–0.929), respectively. The nomogram based on the FIGO stage and Radscore could individually predict the OS probability with good discrimination and calibration. The high-risk patients had shorter OS compared with the low-risk patients (p < 0.05).ConclusionRadiomics has the potential for noninvasive risk stratification and may improve the prediction of OS in patients with cervical cancer when added to the FIGO stage.

EcRD: Edge-Cloud Computing Framework for Smart Road Damage Detection and Warning

Road damages have caused numerous fatalities, thus the study of road damage detection, especially hazardous road damage detection and warning is critical for traffic safety. Existing road damage detection systems mainly process data at cloud, which suffers from a high latency caused by long-distance. Meanwhile, supervised machine learning algorithms are usually used in these systems requiring large precisely labeled data sets to achieve a good performance. In this article, we propose EcRD: an edge-cloud-based road damage detection and warning framework, that leverages the fast-responding advantage of edge and the large storage and computation resources advantages of cloud. There are three main contributions in this article: we first propose a simple yet efficient road segmentation algorithm to enable fast and accurate road area detection. Then, a light-weighted road damage detector is developed based on gray level co-occurrence matrix features at edge for rapid hazardous road damage detection and warning. Furthermore, a multitypes road damage detection model is introduced for long-term road management at cloud, embedded with a novel image generator based on cycle-consistent adversarial networks which automatically generates images with labels to further improve road damage detection accuracy. By comparing with the state-of-the-art, we demonstrate that the proposed EcRD can accurately detect both hazardous road damages at edge and multitypes road damages at cloud. Besides, it is around 579 times faster than cloud-based approaches without affecting users' experience and requiring very low storage and labeling cost.

Descending stairs and floors classification as control reference in autonomous smart wheelchair

Disability is a disruption or limitation of a person’s body functions in carrying out daily activities. A person with physical disabilities needs an assistive device such as a wheelchair. The latest wheelchair development is the smart wheelchair. Smart wheelchairs require a control system to detect obstacles quickly. This aims to provide safety, especially for users. One of the obstacles that are quite dangerous is descending stairs. Therefore the researchers propose a descending stairs detection system for smart wheelchairs. The proposed method in this study is the gray level co-occurrence matrix (GLCM) as the feature extraction algorithm, learning vector quantization (LVQ) as the classification algorithm, and sequential forward selection (SFS) for feature selection. Based on the simulation result, the SFS feature selection gets two selected GLCM features. The best accuracy is 94.5% with the selected features, namely contrast and dissimilarity. This result has an increase in accuracy when compared to using the GLCM 6 feature method with the LVQ classification that does not use feature selection, where the method gets 92.5% accuracy in off-time testing. Accuracy decreased to 78.21% when detecting floors and 89.06% when detecting descending stairs in real-time system testing.

Fish Image Classification by XgBoost Based on Gist and GLCM Features

Classification of fish image is a complex issue in the field of pattern recognition. Fish classification is a complicated task. Physical shape, size, orientation etc. made it complex to classify. Selection of appropriate feature is also a great issue in image classification. Classification of fish image is very important in fishing service and agricultural field, fish industry, survey applications of fisheries and in other related area. For the assessment and counting of fishes, classification of fish image is also necessary as it can save time. This paper presents a fish image classification method with the robust Gist feature and Gray Level Co-occurrence Matrix (GLCM) feature. Noise removal and resizing of image is applied as pre-processing task. Gist and GLCM feature are combined to make a better feature matrix. Features are also tested separately. But combined feature vector performs better than individual. Classification is made on ten types of raw images of fish from two datasets -QUT and F4K dataset. The feature set is trained with different machine learning models. Among them, XgBoost performs with 90.2% and 98.08% accuracy for QUT and F4K dataset respectively.

Breast cancer diagnosis system using hybrid support vector machine-artificial neural network

Breast cancer is the second most common cancer occurring in women. Early detection through mammogram screening can save more women’s lives. However, even senior radiologists may over-diagnose the clinical condition. Machine learning (ML) is the most used technique in the diagnosis of cancer to help reduce human errors. This study is aimed to develop a computer-aided detection (CAD) system using ML for classification purposes. In this work, 80 digital mammograms of normal breasts, 40 of benign and 40 of malignant cases were chosen from the mini MIAS dataset. These images were denoised using median filter after they were segmented to obtain a region of interest (ROI) and enhanced using histogram equalization. This work compared the performance of artificial neural network (ANN), support vector machine (SVM), reduced features of SVM and the hybrid SVM-ANN for classification process using the statistical and gray level co-occurrence matrix (GLCM) features extracted from the enhanced images. It is found that the hybrid SVM-ANN gives the best accuracy of 99.4% and 100% in differentiating normal from abnormal, and benign from malignant cases, respectively. This hybrid SVM-ANN model was deployed in developing the CAD system which showed relatively good accuracy of 98%.

Quantitative assessment of the maxillary sinusitis using computed tomography texture analysis: odontogenic vs non-odontogenic etiology

The purpose of this study was to investigate computed tomography (CT) texture features of mucosal thickening of maxillary sinus mucosa to differentiate odontogenic maxillary sinusitis (OMS) from non-odontogenic maxillary sinusitis (NOMS). Eighteen OMS patients and age- and gender-matched 18 NOMS patients who underwent sinus CT were retrospectively reviewed. OMS patients were identified by histopathological examination of tissues excised at surgery combined with CT imaging findings. Patients with mucosal thickening in the maxillary sinus without apical periodontitis or advanced periodontal bone loss near the maxillary sinus on CT were defined as NOMS. Patients with thin mucosal thickening (< 10mm), cyst, tumor, post-operative deformity, severe metal artifact precluding visualization of the maxillary sinus, and age younger than 20years were excluded. CT texture features of the mucosal thickening were analyzed using an in-house developed Matlab-based texture analysis program. Forty-five texture features were extracted from each segmented volume. The results were tested with the Mann-Whitney U test. Six histogram features (mean, median, standard deviation, entropy, geometric mean, harmonic mean) and two gray-level co-occurrence matrix features (entropy, correlation) showed significant differences between OMS and NOMS patients. CT texture analysis revealed the quantitative differences between OMS and NOMS. The texture features can serve as a quantitative indicator of maxillary sinusitis to differentiate between OMS and NOMS and help prevent incorrect treatment choices.