In this paper, an autonomous brain tumor segmentation and detection model is developed utilizing a convolutional neural network technique that included a local binary pattern and a multilayered support vector machine. The detection and classification of brain tumors are a key feature in order to aid physicians; an intelligent system must be designed with less manual work and more automated operations in mind. The collected images are then processed using image filtering techniques, followed by image intensity normalization, before proceeding to the patch extraction stage, which results in patch extracted images. During feature extraction, the RGB image is converted to a binary image by grayscale conversion via the colormap process, and this process is then completed by the local binary pattern (LBP). To extract feature information, a convolutional network can be utilized, while to detect objects, a multilayered support vector machine (ML-SVM) can be employed. CNN is a popular deep learning algorithm that is utilized in a wide variety of engineering applications. Finally, the classification approach used in this work aids in determining the presence or absence of a brain tumor. To conduct the comparison, the entire work is tested against existing procedures and the proposed approach using critical metrics such as dice similarity coefficient (DSC), Jaccard similarity index (JSI), sensitivity (SE), accuracy (ACC), specificity (SP), and precision (PR).
Medical data classification analysis the medical data of the patients to predict the diseases risk. Data mining techniques were highly used in the medical data classification and predicted the diseases. Many existing methods were use the various classifier and feature selection to improve the performance of the classification. Although data imbalance problem is need to be solved for increases the performance. In this research, Synthetic Minority Over-sampling TEchnique (SMOTE) techniques is used for solving the data imbalance problem and Recurrent Neural Network (RNN) was used for the classification. The SMOTE method based on the k Nearest Neighbor (kNN) for the over-sample and under-sample the attributes. The RNN process the instance independent of the previous instance for the classification. Four medical datasets of University of California, Irvine (UCI) were used to evaluate the effectiveness of the proposed SMOTE-RNN method. The proposed SMOTE-RNN method has the accuracy of 85 % while existing method has 82 % accuracy.
Breast cancer is a very dangerous disease that mainly affects women. It is a deadliest disease that highly affects the women's life. Therefore, it is necessary to predict and classify this deadly disease for early diagnosis. There exist numerous data mining techniques for early prediction and classification of this disease. The big data based analytical model provides the better solution for storing, manipulating, and analyzing a great number of mammographic images. In this article, a new improved fractional rough fuzzy K-means clustering strategy is considered for disease prediction. Then, a new Tunicate Swarm Algorithm (TSA) is introduced to optimize the weight parameters.TSA is a bio-inspired metaheuristic optimization approach. Finally, the labeled ensemble classifier (LEC) is utilized for classifying the stages of breast cancer as malignant and benign. Here, the data is randomly generated from breast cancer Wisconsin dataset (diagnosis) obtainable on UCI machine learning repository. The proposed strategy is compared with different existing strategies, like Logistic Regression Classifier, Random Forest Classifier. From the analysis, it is observed that the proposed big data based analytical model using LEC provides 99.3% accuracy that is very high when compared to the accuracy of existing approaches.
Nowadays, several researchers are facing challenges on the prediction of diseases from the huge volume of medical databases. So, researchers are using data mining techniques like association rules, classification and clustering to address challenges. The physicians make the right decisions for successful diagnosis of various diseases by using the prediction. The existing work classifies the data to predict the certain diseases, but still it faces the difficulties due to overfitting in the training data. The main aim of this research work is to classify the Medical Data (MD) by developing the feature selection based approach in MD Classification as MDC. The irrelevant features are eliminated from the MD by using Recursive Feature Elimination (RFE) method, then ranked the features to reduce the computation cost of the proposed method. The ranked features from the RFE are given as input to the Fuzzy Neural Network (FNN) with Reinforcement Learning (RL), which is used for classification. The proposed RFE-FNN method has the accuracy of the 98.57%, 98.15% sensitivity, 98.64% specificity and 95.47% F-Measure in Heart disease dataset.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.