Deep learning based 3D multimodal CNN for brain tumor segmentation and detection

Brain tumors present a significant challenge to healthcare professionals and can impact individuals of any age. Despite advancements in medicine, early detection and effective treatment remain challenging, often resulting in poor patient outcomes. To address this issue, a convolutional neural network (CNN) model is proposed for segmenting three-dimensional (3D) computed tomography (CT) images to predict the type of brain tumor present. This approach involves applying various pre-processing techniques to the widely used BraTS2020 and BraTS2021 datasets, which are essential in brain tumor research. The pre-processing methods aim to enhance the quality of the CT images, making them suitable for use in a 10-layer CNN architecture for brain tumor classification and prediction. By accurately identifying the type of brain tumor, the most effective treatment strategy can be determined. Extensive experiments on these datasets show that the proposed approach outperforms other contemporary 3D segmentation models, achieving superior recall, precision, Hausdorff, and F1-Score metrics, with Dice Similarity Coefficients of 89.63% and 88.96%, respectively. The proposed method demonstrates significant potential to improve the early detection and treatment of brain tumors, leading to better patient outcomes. Using a CNN model for image segmentation and classification allows healthcare professionals to identify and classify brain tumors accurately and efficiently, enabling the most effective treatment strategies for their patients.