No abstract

Classification of brain neoplasm images is one of the most challenging research areas in the field of medical image processing. The main objective of this study is to design a brain neoplasm classification system that can be trained using multiple various sized MR images of different institutions. The proposed method is a generalized classification system; it can be used in a single institute or in a number of institutions at the same time, without any restriction on image size. The generalization and unbiased capability of the proposed method can bring researchers on a single platform to work on some standard forms of computer aided diagnosis system with more efficient diagnostic capabilities. In this study, a suitable size of moveable rectangular window is used between segmentation and feature extraction stages. A semiautomatic, localized region based active contour method is used for segmentation of brain neoplasm region. Discrete wavelet transform for feature extraction, principal component analysis for feature selection and support vector machine is used as classifier. For the first time MR images of 2 sizes and from different institutions are used in training and testing of brain neoplasm classifier. Three glioma grades were classified using 92 MR images. The proposed method achieved the highest accuracy of 88.26%, the highest sensitivity of 92.23% and the maximum specificity of 93.93%. In addition, the proposed method is computationally less complex, requires shorter processing time and is more efficient in terms of storage capacity.

Brain cancer has remained one of the key causes ofdeaths in people of all ages. One way to survival amongst patientsis to correctly diagnose cancer in its early stages. Recentlymachine learning has become a very important tool in medicalimage classification. Our approach is to examine and comparevarious machine learning classification algorithms that help inbrain tumor classification of Magnetic Resonance (MR) images.We have compared Artificial Neural Network (ANN), K-nearestNeighbor (KNN), Decision Tree (DT), Support Vector Machine(SVM) and Naïve Bayes (NB) classifiers to determine theaccuracy of each classifier and find the best amongst them forclassification of cancerous and noncancerous brain MR images.We have used 86 MR images and extracted a large number offeatures for each image. Since the equal number of images, havebeen used thus there is no suspicion of results being biased. Forour data set the most accurate results were provided by ANN. Itwas found that ANN provides better results for medium to largedatabase of Brain MR Images.

No abstract

For classifying brain tumors with small datasets, the knowledge-based transfer learning (KBTL) approach has performed very well in attaining an optimized classification model. However, its successful implementation is typically affected by different hyperparameters, specifically the learning rate (LR), batch size (BS), and their joint influence. In general, most of the existing research could not achieve the desired performance because the work addressed only one hyperparameter tuning. This study adopted a Cartesian product matrix-based approach, to interpret the effect of both hyperparameters and their interaction on the performance of models. To evaluate their impact, 56 two-tuple hyperparameters from the Cartesian product matrix were used as inputs to perform an extensive exercise, comprising 504 simulations for three cutting-edge architecture-based pre-trained Deep Learning (DL) models, ResNet18, ResNet50, and ResNet101. Additionally, the impact was also assessed by using three well-known optimizers (solvers): SGDM, Adam, and RMSProp. The performance assessment showed that the framework is an efficient framework to attain optimal values of two important hyperparameters (LR and BS) and consequently an optimized model with an accuracy of 99.56%. Further, our results showed that both hyperparameters have a significant impact individually as well as interactively, with a trade-off in between. Further, the evaluation space was extended by using the statistical ANOVA analysis to validate the main findings. \({F}\)—test returned with \({p < 0.05}\), confirming that both hyperparameters not only have a significant impact on the model performance independently, but that there exists an interaction between the hyperparameters for a combination of their levels.