MRI image analysis for cerebrum tumor detection and feature extraction using 2D U-ConvNet and SVM classification

Precise tumor segmentation plays a significant role in radio surgery arrangement and the evaluation of radiotherapy treatment efficiently. To increase the performance and lessen the unpredictability included in the medical image segmentation, transform-based 2D U-ConvNet cerebrum tumor segmentation has been explored. Additionally, to improve the exactness of the support vector machine (SVM) classification, appropriate attributes are extricated from every segmented tissue. The simulation outcomes of this method have been assessed and confirmed for the performance and quality validation of MRI images based on various performance evaluation parameters. This method accomplished the segmentation accuracy of 93.8%, the sensitivity of 87.6%, the specificity of 94.8%, and the recall of 85.6%, exhibiting the efficacy of the proposed method for recognizing ordinary and tumor tissues from MRI images.