Accurate and early detection of cancer is crucial for effective treatment and improved patient outcomes. Machine learning (ML) and deep learning (DL) techniques have demonstrated significant potential in cancer research, enabling precise cancer classification, biomarker identification, and prognostic predictions. This study focuses on the application of 3D convolutional neural networks (CNNs) for the detection of glioblastoma, an aggressive form of brain cancer, from multimodal magnetic resonance imaging (MRI) data. The dataset used comprises MRI scans from 40 patients with glioblastoma, including pre-operative, post-operative, and follow-up time points across different MRI sequences. The performance of individual sequence models and a multi-sequence model combining all available MRI modalities is evaluated and compared based on training accuracy, testing accuracy, training loss, and validation loss. The multi-sequence model achieved the highest testing accuracy, demonstrating the benefits of incorporating diverse imaging information for robust glioblastoma detection.