Search

This research aims to develop a brain tumor detection model by utilizing the machine learning techniques and Convolutional Neural Network (CNN). A significant matter to address is revolving around early detection and the proper handling regarding the brain tumor. This research's methodology consists of collecting the dataset, identifying the tools and language to use, prepare and preprocessing the data, data augmentation, splitting and label encoding, building the model architecture, compiling the model, training, and evaluating the model, predicting the model, and comparing it with other models. Dataset consists of 7022 MRI images, divided into training and testing subsets; and four classes: glioma, meningioma, pituitary, and no tumor. There are four different CNN models that have been built and evaluated, namely VGG16, InceptionV3, ResNet50, and DenseNet121. The result gained shows VGG16 with the best performance achieving an accuracy rate of 96.43%, followed by DenseNet121 (94,96%), InceptionV3 (92,40%), and ResNet50 (78,69%). Although there is still room for improvement regarding overfitting and increasing the models' overall performance, this result is promising enough to enhance early diagnosis and offer an appropriate and effective treatment for patients. [ABSTRACT FROM AUTHOR]

Brain Tumor is the abnormal growth of cancerous cells in the brain. Being able to diagnose them quickly allows practitioners to take preventive measures early, increasing survival rate. Currently, most research utilizes the popular CNN for classification. In this study we aim to find a better classifier by implementing an ensemble method, specifically using CNN and XG-Boost. The dataset we used consists of 3652 MRI images under 2 classes: tumor and no tumor. Our experiment starts off by developing a CNN model. We then used the model for feature extraction and trained an XG-Boost Classifier on the extracted features. The results show an improvement from a 97.61% to a 99.43% testing accuracy going from the CNN to the ensemble model. This demonstrates the potential of ensemble methods for enhancing standard CNN models for tasks like this. This could also open avenues for further advancements in clinical image diagnosis by utilizing the potential of deep learning in combination with ensemble learning. [ABSTRACT FROM AUTHOR]

Brain tumors arise from the emergence of abnormal cells in brain tissue and are considered one of the most perilous conditions affecting individuals of all ages, including both children and adults. The disease advances swiftly, and the likelihood of survival diminishes significantly without prompt and adequate treatment. Hence, accurate diagnosis and meticulous treatment planning play a pivotal role in improving the patient's life expectancy. Neurologists and radiologists play a crucial role in the early detection of brain tumors. However, manually identifying and segmenting brain tumors from Magnetic Resonance Imaging (MRI) data poses significant challenges and is susceptible to inaccuracies. The need for an automated brain tumor detection method becomes imperative to achieve early detection of brain tumors. Objective: The objective of the paper is to measure the capability of You Only Look Once version 5 (YOLOv5x) in brain tumor detection in the early stage so that patients can be treated accordingly. Methods: YOLOv5x is examined; Brats (Brain Tumor Segmentation) image and roboflow dataset has been used. Model performance is evaluated using precision, recall rate, F1 score, and Mean Average Precision (mAP). Results: YOLOv5x exhibited precision (98.7), recall (95.6), F1-score (93), mAP at a learning rate of 0.5 (98.4), and the total time taken for implementation of work is 193.20 minutes. Conclusion: YOLOv5x showed improved performance for the detection of brain tumors on dynamic contrast-enhanced MRI when compared with state of art existing work. It is also the fastest and accurate method indicating a greater potential for clinical application. [ABSTRACT FROM AUTHOR]

• Pre-processing is done by Non-Local Means (NLM) filter. • Image segmentation is done using V-Net. • Brain tumour is achieved by proposed C-SAN. Magnetic Resonance Imaging (MRI) is a medical imaging technique that uses strong magnetic fields and radio waves to generate detailed images of the brain and other organs. Brain tumor is the enlargement of abnormal cells that leads to cancer. The detail about abnormal tissue growth in brain is recognized using MRI. An MRI image is a detailed, three-dimensional image of the inside of the body produced by a non-invasive medical imaging procedure. The development of MRI technology is highly complex and is continually being refined by researchers to equip doctors with enhanced capabilities for patient treatment. Accordingly, this paper proposes a Convolutional stacked Autoencoder Network (C-SAN) for brain tumor detection employing MRI image. First, the acquired image is subjected to preprocessing, which is done by Non-Local Means (NLM) filter. After that, the filtered image is subjected to segmentation phase, which is done by V-Net. The segmented image is then allowed to feature extraction, where features of the image, such as grey level difference statistics, statistical features including coarseness, dissimilarity, autocorrelation and homogeneity, Mean-Variance and Median based Local binary pattern (MVM-LBP) with Discrete Wavelet Transform (DWT), Histogram of Gradients (HOG) and Local Vector Pattern (LVP) are extracted. Finally, brain tumor detection is performed by C-SAN, which is devised by integrating Convolutional Neural Network (CNN) and Deep Stacked Autoencoder (DSAE). The performance of the proposed method is analyzed using metrics, such as accuracy, sensitivity, and specificity. The proposed C-SAN obtained the values of 0.909, 0.958, and 0.928 for accuracy, sensitivity, and specificity, respectively. [ABSTRACT FROM AUTHOR]

• We propose GATE-CNN architecture for binary classification of brain tumor using MR (Magnetic Resonance) images. • We have implemented our model on three different datasets; all consists of MRI with a significant difference in number of images. • The first dataset consists of tumorous and non-tumorous MRIs, second consists of MRIs with glioma and pitutiary (types of brain tumor), and the third consists of cancerous and non-cancerous (severity of brain tumor) images. • For all three datasets we found better classification results than any of the existing state-of-the-art models. Early and accurate detection is a solitary precaution to overcome the brain tumor. Otherwise, it will result in a deadly disease. Brain tumor (BT) detection using magnetic resonance is an essential and challenging job in the medical domain. To combat the aggressive spreading of BT, clinical imaging (MRI and X-Ray) can be an appropriate method for diagnosis. In this paper, we applied attention based image classification, which is a new paradigm for obtaining improved accuracy than the state-of-the-art models. We propose a GATE-CNN (Graph Attention AutoEncoder-Convolution Neural Network) model in this work for the classification of BT. We calculate the attention values of the neighboring pixels on each and every pixel present in the graph then process the graph using GATE framework and the processed graph with attention values is then passed to CNN framework for generation of final output. Further, we apply Adamax optimizer to optimize the training hyperparameters of CNN. We perform the BT classification using the proposed method on three different datasets among benign and malignant BT for first dataset, glioma and pituitary for second dataset, and normal and abnormal BT images for third dataset; all the three datasets consists of MRI images. We then compare the proposed model with a variety of CNN (Deep-CNN, Multi-Input CNN, and basic CNN) models dealing with several types of medical imaging (thermal image, MRI, X-Ray, etc.). Our model results in the accuracy of 98.27% for first dataset, 99.83% for second dataset, and 98.78% for third dataset demonstrating preferable network performance than the already established state-of-the-art CNN models. [ABSTRACT FROM AUTHOR]

• To design a brain tumor detection and classification with deep learning methods. • To develop a Hybrid Beluga Whale Glowworm Swarm Optimization (HBWGSO) algorithm. • To tune the epochs, learning rate in MDIN and hidden neuron count in RAN by HBWGSO. • To design a new HBWGSO algorithm for attaining the effective segmented regions. • To implement a Residual Attention Multiscale Dilated Inception Network (MDIN). A tumor is formed by controlled and rapid cell growth in the brain. If it is not treated in the early stages, it leads to death. Despite many important efforts and promising solutions, accurate segmentation and classification remain a challenge. Traditional automated models have complex architectures, high computing systems, and large amounts of data. Also, most of the existing models still rely on manual intervention. To address all the limitations, a new deep-learning approach is proposed. Initially, brain images are collected from standard sources and passed to the pre-processing stage. In this stage, the input images are pre-processed using scaling, contrast enhancement, and Anisotropic Diffusion Filtering (ADF). Later, the resultant images are provided to the image segmentation stage. Here, image segmentation is performed using adaptive transunet3+, and also their parameters are optimized by using the Hybrid Beluga Whale Glowworm Swarm Optimization (HBWGSO). Further, brain tumor classifications are performed with the hybridization of Residual Attention Network (RAN) and Multiscale Dilated Inception Network (MDIN) termed RA-MDIN, and the model parameters are optimally selected by using the designed HBWGSO approach. Through the experimentation results, the proposed model tends to provide effective classification results. Thus, the recommended system guarantees to yield relatively satisfactory outcomes over conventional mechanisms. [ABSTRACT FROM AUTHOR]

• To develop a novel early-stage brain tumor prediction model using ensemble deep networks with enhanced optimization for forecasting the brain tumors in the patients to treat them with the right medications, which helps to save their lives. • To choose the optimal features, where the weight of the features is optimized by the designed Enhanced Artificial Rabbits Optimizer (EARO) algorithm for attaining the best weight value by maximizing the correlation coefficient among divergent classes and minimizing the correlation value among similar classes. • To design an EARO algorithm for optimizing the weights during the feature selection process to increase the performance over brain tumor prediction. • To ensemble 1-Dimensional Convolutional Neural Network (1DCNN), Bidirectional Long Short-Term Memory (Bi-LSTM), and Deep Temporal Convolution Network (DTCN) using high ranking method along with the dilation and attention network for designing Dilated and Attention-based Ensemble Learning Network (DAELNet) for predicting the brain tumors effectively in the prior stage. • To validate the performance of the developed early-stage brain tumour prediction model over the existing models and algorithms concerned with different effectiveness measures. The integration of deep learning into brain data analysis has notably boosted the field of biomedical data analysis. In the context of intricate conditions like cancer, various data modalities can reveal distinct disease characteristics. Brain data has the potential to expose additional insights compared to using the data sources in isolation. Moreover, techniques are selected and prioritized based on the speed and accuracy of the data. Therefore, a new deep learning technique is assisted in predicting the brain tumor from the brain data to provide accurate prediction outcomes. The brain data required for predicting the brain tumor is garnered through various online sources. Then, the collected data are applied to the data preprocessing phase for cleaning the collected brain data and then applied to the data transformation method to improve the efficiency for providing better decision-making over prediction. The transformed data is then offered to the weighted feature selection process, where the weights of the features are optimized through the proposed Enhanced Artificial Rabbits Optimizer. The selection of weighted features is primarily adopted for solving the data dimensionality issues and these resultant features are given to the Dilated and Attention-based Ensemble Learning Network to provide the effective prediction outcome, where the deep learning structures like 1-Dimensional Convolutional Neural Networks, Bidirectional Long Short-Term Memory (BiLSTM), Deep Temporal Convolution Network are ensembled in the DAEL network. Finally, the prediction outcome attained from the proposed model is validated through the existing brain tumor prediction frameworks to ensure the efficacy of the implemented scheme. [ABSTRACT FROM AUTHOR]

This study introduces a circular ring and square metasurfaces-based sensor employing graphene and gold layers for enhanced brain tumor detection. The sensor architecture includes dual graphene circular metasurfaces and a gold square resonator layer on a silicon dioxide substrate. A comprehensive parametric investigation explores the influence of geometric parameters and graphene chemical potential on sensor performance. The sensor achieves a remarkable sensitivity of 769 GHzRIU−1, a high-quality factor of 6.514, a low detection limit of 0.182 RIU, and an FOM of 1.148RIU−1. Electric field intensity analysis identifies optimal transmission frequencies. By employing the XGBoost regressor, the analysis significantly reduces simulation time and resource consumption by at least 80 %. Simulation validation confirms that despite 80 % reduction in simulation time, high prediction accuracy remains achievable, with an impressive R2 score ranging from 0.9 to 1. The proposed sensor presents a promising non-invasive approach for early brain tumor diagnosis, with potential applications in biomedical imaging and other related fields. [Display omitted] [ABSTRACT FROM AUTHOR]

Brain tumor detection has long been a significant and challenging issue in medical research. In this study, we propose an unsupervised pseudo-deep method for detecting brain tumors in Magnetic Resonance (MR) images. Our approach utilizes iterative spectral Co-Clustering and Fuzzy C-means techniques to achieve precise segmentation results. In each iteration, the algorithm recognizes the tumor block from its input image using spectral Co-Clustering. The selected block from the previous iteration serves as the input for the subsequent iteration. In the final iteration, Fuzzy C-Means is applied to the last selected block to extract the tumor, and a scheme is proposed to determine the location of the tumor in the original image. Through this iterative pseudo-deep method, our approach exhibits a layered-like structure, resembling deep learning architectures, within the context of unsupervised methods. We evaluated our approach using the BraTS2020 and BraTS2021 datasets and assessed its performance using key metrics. Our method achieved values of 99.12% and 81.42% for accuracy and dice coefficient on BraTS2020 and also 99.21% and 82.03% on the BraTS2021 dataset, surpassing the performance of existing unsupervised methods reported in the literature. Moreover, our approach demonstrates notably superior performance when dealing with complex images. The proposed method offers a unique perspective in the application of unsupervised techniques for brain tumor detection using a pseudo-deep structure. [ABSTRACT FROM AUTHOR]

The detection of Brain cancer is an essential process, which is based on the clinician's knowledge and experience. An automatic tumor classification model is important to handle radiologists to detect the brain tumors. However, the precision of present model should be enhanced for appropriate treatments. Numerous computer-aided diagnosis (CAD) models are offered in the literary works of medical imaging to help radiologists concerning their patients. This paper proposes an optimization-driven technique, namely Whale Harris Hawks optimization (WHHO) for brain tumor detection using MR images. Here, segmentation is performed using cellular automata and rough set theory. In addition, the features are extracted from the segments, which include tumor size, Local Optical Oriented Pattern (LOOP), Mean, Variance, and Kurtosis. In addition, the brain tumor detection is carried out using deep convolutional neural network (DeepCNN), wherein the training is performed using proposed WHHO. The proposed WHHO is designed by integrating Whale optimization algorithm (WOA) and Harris hawks optimization (HHO) algorithm. The proposed WHHO-based DeepCNN outperformed other methods with maximal accuracy of 0.816, maximal specificity of 0.791, and maximal sensitivity of 0.974, respectively. [ABSTRACT FROM AUTHOR]

The central organ of the human nervous system is the brain, which receives and sends stimuli to the various parts of the body to engage in daily activities. Uncontrolled growth of brain cells can result in tumors which affect the normal functions of healthy brain cells. An automatic reliable technique for detecting tumors is imperative to assist medical practitioners in the timely diagnosis of patients. Although machine learning models are being used, with minimal data availability to train, development of low-order based models integrated with machine learning are a tool for reliable detection. In this study, we focus on comparing a low-order model such as proper orthogonal decomposition (POD) coupled with convolutional neural network (CNN) on 2D images from magnetic resonance imaging (MRI) scans to effectively identify brain tumors. The explainability of the coupled POD-CNN prediction output as well as the state-of-the-art pre-trained transfer learning models such as MobileNetV2, Inception-v3, ResNet101, and VGG-19 were explored. The results showed that CNN predicted tumors with an accuracy of 99.21% whereas POD-CNN performed better with about 1/3rd of computational time at an accuracy of 95.88%. Explainable AI with SHAP showed MobileNetV2 has better prediction in identifying the tumor boundaries. Integration of POD with CNN is carried for the first time to detect brain tumor detection with minimal MRI scan data. This study facilitates low-model approaches in machine learning to improve the accuracy and performance of tumor detection. • Proper Orthogonal Decomposition (POD), a reduced order model, is employed for brain tumor detection utilizing 2D MRI images. • Pre-trained transfer learning models, such as MobileNetV2, achieving an accuracy of 99.21%. • POD, paired with CNN, spotted tumors in 1/3 of the time of MobileNetV2, with 95.88% accuracy. [ABSTRACT FROM AUTHOR]

The main goal of this paper is to attain accurate and automatic detection of brain tumors in gray magnetic resonance images using reducing the local dimensions. We propose a novel model called Improved Mixtures of Factor Analyzers based on Dynamic Co-Clustering (IMFADCC) for the detection and localization of brain tumors. After image preprocessing and enhancement, the optimal numbers of row clusters and column clusters corresponding to each row cluster, alongside other model parameters, are determined concurrently based on the size of the tumor by our model. Then, using the optimal values obtained, the image is co-clustered by simultaneously clustering rows and columns until the block containing the tumor is identified. The output image is divided into a certain number of blocks, one of which contains the tumor. For post-processing, the identified block is binarized using minimum error thresholding. The proposed model accurately detects and locates the block binarized within the original image while considering the remainder of the image as background. The effectiveness of the proposed method was assessed by evaluating its accuracy, Dice, intersection on union, geometric mean, receiver operating characteristic curve, specificity, and sensitivity criteria on the BraTS2018, BraTS2019, and BraTS2020 datasets. The results show that our method has significant performance and is highly accurate in detecting tumors of different sizes in complex and high-size images. Moreover, our method outperforms the existing diagnostic methods. [ABSTRACT FROM AUTHOR]

• In nervous system, the uncontrolled growth of cells leads to cause brain tumors. • The CSA-MLP is proposed to perform the classification of tumor cells accurately. • To perform an exact segmentation of normal and abnormal brain cells. • Accuracy, precision, and F1-score metrics are used to validate the efficiency. The Brain tumor is considered an unusual growth of cells in the nervous system that restricts the normal functionality of the brain. However, is generated in the skull and pressures the brain which affects the health of a person. So it is essential to detect and classify the brain tumor at an early stage before reaching the severity level. Meanwhile, brain tumor detection is performed based on MRI images which are considered an effective diagnosis system. But the detection and classification using MRI images is obtained as a complex task and cannot show the difference between normal and abnormal cells. So to overcome this issue the Crossover Smell Agent Optimized Multilayer Perception (CSA-MLP) is proposed to perform the exact detection and classification of tumor cells from MRI images. The images are collected from three datasets namely MR, Brain MRI, and Brain tumor datasets and they are preprocessed to remove the unwanted noise. After preprocessing the features of the images are extracted to perform the classification process. Moreover, the Convolutional Neural Network (CNN) classifier is used to classify healthy and unhealthy brain cells. The Multi-Layer Perceptron (MLP) is employed for the classification category that minimized the errors and enhanced the performance of the proposed method. The MLP is integrated with the CSA optimization algorithm to improve classification accuracy. The experimentation results revealed that the proposed method achieved a better accuracy of about 98.56% which enhanced the effectiveness compared to existing methods. [ABSTRACT FROM AUTHOR]

The brain tumor is one of the life-threatening diseases. It has been classified into the malignant and benign brain tumor. Benign tumor cells have very similar characteristics to its surrounding healthy cells. Hence, its accurate detection and segmentation are one of the difficult and challenging tasks. Existing state-of-the-art methods used for detection and segmentation of such tumor cells are unable to locate such benign tumor region. In this manuscript, a novel fractional partial differential equation (FPDE) based super-diffusive models is being proposed. The fractional nature enables to detect and segment such low variational of benign brain tumor region more accurately. It is also found that these partial differential models are computationally inefficient due to dependence on the mesh. Hence, the FPDE is being solved using the mesh-free method. Both qualitative and quantitative evaluation has been done. It is found that the proposed model is superior in benign brain tumor detection and segmentation. [ABSTRACT FROM AUTHOR]

The advent of attention-based architectures in medical imaging has ushered in an era of precision diagnostics, particularly in the detection and classification of brain tumors. This study introduced an innovative knowledge distillation framework employing a tripartite attention mechanism within transformer encoder models, specifically tailored for the identification of multiple brain tumor classes through magnetic resonance imaging (MRI). The proposed methodology synergistically harnesses the capabilities of large, highly parameterized teacher models to train more compact, efficient student models suitable for deployment in resource-constrained environments such as the internet of medical things and smart healthcare devices. Utilizing a diverse array of MRI sequences—including T1, contrast-enhanced T1, and T2—this study accounts for the nuanced variations across brain tumor classes derived from three extensive datasets. The tripartite attention mechanism addresses the limitation of traditional attention models by innovatively integrating temperature-softening neighborhood attention, global attention, and cross-attention layers. This sophisticated approach allows for a richer and more nuanced feature representation, capturing both local and global contextual information and intricate tumor features within MRI scans. This is supplemented by a unique augmentation pipeline and shifted patch tokenization technique, which enrich the model's input representation, especially for underrepresented classes. Through meticulous experimentation and ablation studies, the study demonstrates that the proposed model not only retains the robustness of its larger counterparts but also delivers enhanced performance metrics. When juxtaposed with benchmarking models—including traditional deep CNNs and various transformer-based architectures—the proposed model consistently showcases superior results. Its effectiveness is reflected in its lower teacher and student losses, commendable Brier scores, and noteworthy top-1 and top-5 accuracies, as well as AUC metrics across all datasets. This paper not only validates the efficacy of knowledge distillation in complex medical image analysis tasks but also provides a promising pathway for the integration of cutting-edge AI techniques in real-world clinical applications, potentially revolutionizing the early detection and treatment of brain tumors. [ABSTRACT FROM AUTHOR]

The aim of this paper is to present a methodology for brain tumor temperature effect extraction from Magnetic Resonance Imaging (MRI). MRI is a medical imaging modality that gives access to a large number of information on the observed tissues and tumors according to the nuclear magnetic resonance parameters. However, some of these parameters may vary depending on the temperature. Tumor cells generate more heat than surrounding healthy cells, which causes a significant changes on MR parameters, the studied MR parameter in this work is spin-lattice relaxation time T1. The problem is expressed mathematically, and computer simulated T1-weighted images are obtained using Spin Echo sequence (SE). The head modeled as a three concentric spheres including spherical tumor. The temperature distribution is calculated using Pennes BioHeat Transfer Equation (PBHTE), it’s solved using Finite Difference Method (FDM). Results show that, heat generation by tumor reduces significantly the T1-weighted signal intensity, which makes brain tumor detection more difficult. The extraction of brain tumor temperature effect from MRI images makes the detection of tumor more accurate. The effect of tumor size on signal intensity also has been presented. [ABSTRACT FROM AUTHOR]

The aim of this paper is to present a GPU parallel algorithm for brain tumor detection to estimate its size and location from surface temperature distribution obtained by thermography. The normal brain tissue is modeled as a rectangular cube including spherical tumor. The temperature distribution is calculated using forward three dimensional Pennes bioheat transfer equation, it's solved using massively parallel Finite Difference Method (FDM) and implemented on Graphics Processing Unit (GPU). Genetic Algorithm (GA) was used to solve the inverse problem and estimate the tumor size and location by minimizing an objective function involving measured temperature on the surface to those obtained by numerical simulation. The parallel implementation of Finite Difference Method reduces significantly the time of bioheat transfer and greatly accelerates the inverse identification of brain tumor thermophysical and geometrical properties. Experimental results show significant gains in the computational speed on GPU and achieve a speedup of around 41 compared to the CPU. The analysis performance of the estimation based on tumor size inside brain tissue also presented. [ABSTRACT FROM AUTHOR]

Brain tumor detection is a very important and challenging task. An efficient detection algorithm is of great importance to the practice of brain tumor medicine. In this paper, we propose a novel parallel optimization algorithm based on a shared-memory environment to solve the SVM classifier for brain tumor detection. Firstly, the HOG algorithm is used to extract brain tumor MR image features and compare them with the wavelet transform method. Secondly, SVM with ℓ 1 -norm loss function is utilized as a classifier. Due to its sparsity, the detection speed is significantly faster. Finally, SMP-SGD, SMP-Momentum, SMP-Adagrad, and SMP-Adam algorithms are proposed and applied to the classifier solution. The experimental results show that the HOG algorithm extracts brain tumor MRI features more effectively than the discrete wavelet transform method. The proposed SMP-SGD algorithm and its variants achieved state-of-the-art accuracy and efficiency for brain tumor detection. [ABSTRACT FROM AUTHOR]

Classification is the main work while detecting the brain tumor images. Prior researchers planned to determine the brain tumors via various classification approaches. But, the error rate and detection time of tumor are high. Therefore, brain tumor detection is improved via classification using the Deep Belief Network Assisted Quadratic Logit BoostClassifier (DBNQLBC) technique. The proposed DBNQLBC technique is employed for increasing the accuracy with lesser error and time using classifying the brain images. The features from the input brain MRI images are taken as input in the DBNQLBC technique to carry out the brain tumor detection. DBNQLBC technique comprises the different types of layers namely input layer, hidden layers, and output layer. From the brain MRI images, the input layer gets the features and it is sent to the hidden layer. In the hidden layer, a quadratic logit boostclassifier is applied to classify the extracted features. Boosting classifier uses the quadratic classifier as a sub-classifier to detect the brain tumor through the likelihood measure. The results of sub-classifiers are merged and create a robust one through diminishing logit loss. The Boosting classifier determines the best classification results that provide higher accuracy results. As a result, input MRI images are accurately categorized into normal and abnormal and the outcomes are displayed at the output layer. From this, brain tumor detection is achieved with lower error, time and higher accuracy. Simulation evaluation is carried out using the metrics such as brain tumor detection accuracy, brain tumor detection time, and false-positive rate based on the number of MRI images. The obtained outcomes ensure the presented DBNQLBC technique effectively increases the brain tumor detection accuracy and reduces the time requirement and false-positive rate than the other methods. [ABSTRACT FROM AUTHOR]

The objective of multilevel thresholding is to segmenting a gray-level image into several distinct homogeneous regions. This paper presents an alternative approach for unsupervised segmentation of natural and medical images to improve the separation between objects in the framework of multi-objective optimization. In contrast to the existing single-objective optimization and entropy-based methods, a multi-objective framework is adopted by combining two objectives based on the Minimum Cross Entropy (MCE) and Rényi Entropy (RE). One of the most competitive Multi-Objective Evolutionary Algorithms (MOEAs) of current interest, called MOEA/D-DE (Decomposition based MOEA with Differential Evolution) is then applied to determine the set of Pareto optimal solutions for these two objectives. The threshold values for multi-level segmentations are obtained from the approximated Pareto Fronts (PFs) generated by MOEA/D-DE. The performance of MOEA/D-DE is also investigated extensively through comparison with other popular nature-inspired single-objective and multi-objective optimizers. Moreover, outcome of the proposed method is evaluated by comparing against the results of other well cited algorithms both qualitatively and quantitatively on test-suites comprising well-known natural and medical test images in order to showcase the efficiency of the proposed algorithm. [ABSTRACT FROM AUTHOR]

Researchers are increasingly interested in leveraging the Internet of Things in medical and healthcare systems to provide better solutions such as remote health monitoring, personal fitness, and chronic disease management. A brain tumor is a potentially fatal cancer caused by the uncontrollable growth of brain cells that affects human blood cells and nerves. However, accurate classification of brain tumors is difficult due to the vastly different anatomical structures of healthy and tumorous tissues. We propose a framework for brain tumor localization and classification based on CenterNet. The proposed method uses the ResNet34 model with an attention block as a base network, which improves feature representation capacity by focusing on tumor locations and aids in tumor classification, particularly for a small tumor. Our method achieves 98.98% accuracy overall. Both qualitative and quantitative analysis demonstrated the efficacy of our approach for accurate detection and classification of the brain tumor than existing latest approaches. [ABSTRACT FROM AUTHOR]

Brain tumor detection received much attention due to its clinical significance for early treatment. Accurate diagnosis and classification of brain tumors are still challenging despite many major contributions are available. Existing methods mainly focus on accuracy in which classification problems like overfitting and underfitting have remained a major concern. This paper presents a technique for brain tumor identification using a deep autoencoder based on spectral data augmentation. In the first step, the morphological cropping process is applied to the original brain images to reduce noise and resize the images. Then Discrete Wavelet Transform (DWT) is used to solve the data-space problem with feature reduction. Finally, a dense layer is proposed for appropriate feature extraction and classification of the brain tumor images. The comparative analysis shows that the proposed algorithm outperforms with an accuracy of 97% and an AUC ROC score of 99.46%. © 2017 Elsevier Inc. All rights reserved. [Display omitted] [ABSTRACT FROM AUTHOR]

• A robust technique is presented for MR image-based brain tumor detection. • The residual blocks improve the classification achievement. • Deep features are extracted from FC layers and residual-convolutional layers. • The L1NSR feature selection algorithm improves the classification performance. One of the most dangerous diseases in the world is brain tumors. After the brain tumor destroys healthy tissues in the brain, it multiplies abnormally, causing an increase in the internal pressure in the skull. If this condition is not diagnosed early, it can lead to death. Magnetic Resonance Imaging (MRI) is a diagnostic method frequently used in soft tissues with successful results. This study presented a new deep learning-based approach, which automatically detects brain tumors using Magnetic Resonance (MR) images. Convolutional and fully connected layers of a new Residual-CNN (R-CNN) model trained from scratch were used to extract deep features from MR images. The representation power of the deep feature set was increased with the features extracted from all convolutional layers. Among the deep features extracted, the 100 features with the highest distinctiveness were selected with a new multi-level feature selection algorithm named L1NSR. The best performance in the classification stage was obtained by using the SVM algorithm with the Gaussian kernel. The proposed approach was evaluated on two separate data sets composed of 2-class (healthy and tumor) and 4-class (glioma tumor, meningioma tumor, pituitary tumor, and healthy) datasets. Besides, the proposed approach was compared with other state-of-the-art approaches using the respective datasets. The best classification accuracies for 2-class and 4-class datasets were 98.8% and 96.6%, respectively. [ABSTRACT FROM AUTHOR]

Abstract: A significant medical informatics task is indexing patient databases according to size, location, and other characteristics of brain tumors and edemas, possibly based on magnetic resonance (MR) imagery. This requires segmenting tumors and edemas within images from different MR modalities. To date, automated brain tumor or edema segmentation from MR modalities remains a challenging, computationally intensive task. In this paper, we propose a novel automated, fast, and approximate segmentation technique. The input is a patient study consisting of a set of MR slices, and its output is a subset of the slices that include axis-parallel boxes that circumscribe the tumors. Our approach is based on an unsupervised change detection method that searches for the most dissimilar region (axis-parallel bounding boxes) between the left and the right halves of a brain in an axial view MR slice. This change detection process uses a novel score function based on Bhattacharya coefficient computed with gray level intensity histograms. We prove that this score function admits a very fast (linear in image height and width) search to locate the bounding box. The average dice coefficients for localizing brain tumors and edemas, over ten patient studies, are 0.57 and 0.52, respectively, which significantly exceeds the scores for two other competitive region-based bounding box techniques. [Copyright &y& Elsevier]

Abstract: In this study, a novel method is proposed for the detection of tumor in magnetic resonance (MR) brain images. The performance of the novel method is investigated on one phantom and 20 original MR brain images with tumor and 50 normal (healthy) MR brain images. Before the segmentation process, 2D continuous wavelet transform (CWT) is applied to reveal the characteristics of tissues in MR head images. Then, each MR image is segmented into seven classes (six head tissues and the background) by using the incremental supervised neural network (ISNN) and the wavelet-bands. After the segmentation process, the head is extracted from the background by simply discarding the background pixels. Symmetry axis of the head in the MR image is determined by using moment properties. Asymmetry is analyzed by using the Zernike moments of each of six tissues segmented in the head: two vectors are individually formed for the left and right hand sides of the symmetry axis on the sagittal plane by using the Zernike moments of the segmented tissues in the head. Presence of asymmetry and the tumors are inquired by considering the distance between these two vectors. The performance of the proposed method is further investigated by moving the location of the tumor and by modifying its size in the phantom image. It is observed that tumor detection is successfully realized for the tumorous 20 MR brain images. [Copyright &y& Elsevier]

The proposed intelligent method for accurate brain tumor detection is composed of five phases, "(a) skull stripping (b) tumor segmentation (c) feature extraction (d) optimal feature selection and (e) classification". Initially, the input MRI image is collected from the Kaggle dataset. This MRI image is subjected to the first step called Pre-processing, which is done using the skull stripping and the entropy-based trilateral filtering methods. The skull stripping removes the non-cerebral tissues, and only the skull part is considered. The skull stripping is done using the Otsu thresholding. The entropy-based trilateral filtering smoothens the images by preserving the edges using a nonlinear combination of nearby pixel values. The graphicalarchitectural diagram of a proposed intelligent method for brain tumor detection has been shown in Graphical abstract. The pre-processed image is given to the tumor segmentation phase. Here the tumors are segmented from the image using fuzzy centroid-based region growing. Once the tumor segmentation is performed, it undergoes the feature extraction phase to extract the features. This is accomplished using four features such as GLCM, GLRM, statistical features, and shape features. Here, the statistical features extract the features from the first-order statistics such as, "mean, median, standard deviation, and variance", and higher-order statistics such as, "RMS, smoothness, kurtosis, and skewness". Similarly, the shape features extract the "area, eccentricity, extent, circularity, and Euler number" as features. For minimizing the complexity that occurs during the classification phase, it is mandatory to choose the most significant features from the long feature length. This is done in the optimal feature selection phase, in which the optimization is done by the hybrid meta-heuristic algorithm by integrating GSO and MVO. The resultant algorithm is termed the GS-MVO. These optimally selected features are subjected to the deep learning algorithm called DBN in the classification phase. An improvement is made in the DBN by optimizing its weight using the same GS-MVO. [Display omitted] • Undergoes pre-processing by the skull stripping and the entropy-based trilateral filtering methods for removing the non-cerebral tissues by the Otsu thresholding and smooth the images by preserving the edges. • Performs the tumor segmentation for separating the tumor from the normal brain tissues using the fuzzy centroid-based region growing. • Extracts the GLCM, GLRM, statistical, and the shape features from the tumor segmented image to more manageable groups for the processing. • Performs the optimal feature selection for choosing the most significant features to minimize the classification complexity by the proposed GS-MVO. • Classifies the MRI brain image as normal or abnormal by the deep learning algorithm called DBN, where the weight is optimized by the same proposed GS-MVO. Nowadays, a Magnetic Resonance Image (MRI) scan acts as an efficient tool for efficiently detecting the abnormal tissues present in the brain. It is a complex process for radiologists to diagnose as well as classify the tumor from several images. This paper develops an intelligent method for the accurate detection of brain tumors. Initially, the pre-processing is performed for the input MRI image using the skull stripping and the entropy-based trilateral filtering methods. Further, fuzzy centroid-based region growing is adopted for segmenting the tumor from the image. Once the tumor is segmented, feature extraction is done using four sets of well-performing features like Gray-Level Co-occurrence Matrix (GLCM), Gray Level Run-length Matrix (GLRM), statistical features, and shape features. The optimal feature selection is performed by the hybrid meta -heuristic algorithm termed Group Search-based Multi-Verse optimization (GS-MVO). Finally, the optimally selected features are given to a deep learning algorithm called Deep Belief Network (DBN). The weight is optimized by the same GS-MVO that classifies the final image as normal or abnormal. The simulation outcomes are performed by the standard benchmark database which proves that the developed technique obtains a high classification accuracy. From the analysis, the accuracy of the proposed GS-MVO-DBN is 9.09% superior to SVM, 7.14% superior to NN, 3.45% superior to DBN, 17.65% superior to CNN, 15.38% superior to NN-CNN, and 1.69% superior to COR-CSO-CNN-NN. The proposed GS-MVO-DBN is very effective in accurately detecting brain tumors. In the future, it is encouraged to work on challenging parts of the tumor region like edema, necrosis, and active regions with the help of the fusion process of multi-modality MRI images and effective pre-processing techniques incorporated with innovative deep learning methods. [ABSTRACT FROM AUTHOR]

• UnetResNext-50: A hybrid Deep CNN model designed for brain tumor detection. • Automatic segmentation and classification of abnormal tumors from Brain MRI. • Publically available datasets with 3929 MR images to develop the model. • Excellent DICE score of 95.73%. • Very efficient with exceptional accuracy of 99.7%. The automatic and accurate detection and segmentation of brain tumors is a very tedious and challenging task for medical experts and radiologists. This paper proposes a hybrid deep convolutional neural network (CNN) model with a large number of layers and parameters for the automatic and accurate prediction and segmentation of brain tumors from the magnetic resonance imaging (MRI) images. The proposed model has a skip connection with cardinality which solves the problem of gradient degradation and also reduces the computational cost of Deep CNN architecture and it also improves the pixel quality at the decoder side. MRI dataset containing a total of 3929 MR images including 1373 images with tumors and 2556 images of normal type (without tumor). The dataset is preprocessed and augmented with 21 parameters before feeding the train images to the proposed models for learning. Performance metrics used for the evaluation of model efficiency are Jaccard Index, DICE score, F1-score, accuracy, precision, and recall. Our model performance is also evaluated by comparing it with the other two models UnetResNet-50 and Vanilla Unet and also with state-of-art techniques. In the post-processing stage, the scores of segmented tumor areas are calculated based on the scores of IoU and DICE and are also presented for comparison with the original images. Performance evaluation metrics show that the proposed model UnetResNext-50 shows excellent efficacy with 99.7% accuracy and a 95.73% DICE score. [ABSTRACT FROM AUTHOR]

• The proposed work aims to explore the use of OSH method in brain tumor detection. • The proposed method combines an off-axis optical scan and an MRI display by a SLM. • The underlying physics of OSH-ACM is the use of in-phase component of the out current. • The famous brain tumors database of BRATS is used to test the proposed system. • Parameters reverted by the optical process are used to investigate the performance. Tumor segmentation from magnetic resonance images (MRI) is an error-prone and time-absorbing process. Recently, optical methods have opened a new avenue to tack with the aforementioned problem. In this paper, we propose a novel architecture adapting the Optical Scanning Holography (OSH) to the detection of the abnormal tissue regions in MRI. The proposed method combines an off-axis optical scan, performed by a heterodyne fringe pattern, and a MR image display ensured by a spatial light modulator. The output in-phase component of the scanned current is collected digitally. Hence, a high-precision distribution of biological tissues is extracted using this in-phase component. Its maximum position is exactly the one of the tumor. Meanwhile, this position is used in an Active Contour Model (ACM) to perform a fast segmentation of the extent corresponding to the tumors. Several images of brain tumors from BRATS database, with tumors having different contrast and form, are used to test the proposed system. Parameters reverted by the optical process are used to investigate the detection performance. Further, in terms of tumor segmentation, the proposed OSH-ACM process has high performance metrics compared to some of recently published method. The underlying physics of the precision superiority, presented by the OSH-ACM, is the high-precision extraction of the abnormal tissue regions by the in-phase component of the scanned current. [ABSTRACT FROM AUTHOR]