Brain tumors (BT) remain a major clinical challenge due to their heterogeneous structure, complex morphology, and limited therapeutic efficacy caused by variability in drug penetration across tumor sub-regions. Accurate characterization of tumor tissues is therefore essential not only for diagnosis but also for optimizing targeted drug delivery, treatment planning, and therapeutic decision-making. In this study, an AI-assisted MRI-based brain tumor characterization framework, termed BTC-MSTCNN-DSVM-MRI, is proposed to support pharmaceutical and therapeutic planning. MRI images obtained from the BRATS 2020 dataset are pre-processed using a Switching Model Stein Variational Sampling Filter (SMSVSF) to perform resizing, normalization, and noise reduction. Data augmentation techniques, including rotation, flipping, and scaling, are applied to enhance dataset diversity and model generalization. A Multi-Anchor Space-Aware Temporal Convolutional Neural Network (MSTCNN) is employed to extract intensity-based, shape-based, and texture features from MRI images, capturing spatial and temporal tumor heterogeneity. These features are classified using Distributed Support Vector Machines (DSVM) to accurately characterize tumor sub-regions into GD-enhancing tumor, peritumoral edema, necrotic tissue, and non-enhancing tumor core, which are clinically relevant for drug delivery optimization. Furthermore, a Super Cell Thunderstorm Algorithm (STA) is utilized to optimize model parameters, improving classification reliability and computational efficiency. Experimental results demonstrate superior performance with an accuracy of 99.55%, precision of 98.75%, and recall of 98.65%, outperforming existing state-of-the-art methods. The proposed framework provides a robust decision-support tool for targeted drug delivery and personalized brain tumor therapy.
Keywords: Brain tumor; MRI; Drug delivery; Therapeutic planning; Deep learning; Pharmaceutical decision support
How to cite this article: Sankaranarayanan R, Susee SK, Senthil Kumar M, Chidhambara Rajan B, Advanced Brain Tumor Classification Utilizing Multi-Anchor Space-Aware Temporal Convolutional Neural Network Integrated With Distributed Support Vector Machines For Accurate MRI Image Analysis. Int J Drug Deliv Technol. 2026;16(2s): 990-1008; DOI: 10.25258/ijddt.16.990-1008