Deep Fusion Model for Brain Tumor Classification Using Fine-Grained Gradient Preservation

TL;DR

This paper tackles brain tumor classification with a resource-efficient, accurate approach suitable for edge deployment. It introduces a fusion architecture that combines ResNet152V2 and a modified VGG16 with gradient-preserving tweaks, augmented by a dual attention mechanism, and uses XGBoost as the final classifier. Quantized to 8-bit and yielding only about 2.8 million trainable parameters, the model achieves 98%+ accuracy on Figshare and Kaggle datasets, with Grad-CAM applied for interpretability. The work demonstrates strong performance improvements over existing approaches while addressing deployment constraints in underdeveloped regions, and discusses future directions in domain adaptation and ablation studies.

Abstract

Brain tumors are one of the most common diseases that lead to early death if not diagnosed at an early stage. Traditional diagnostic approaches are extremely time-consuming and prone to errors. In this context, computer vision-based approaches have emerged as an effective tool for accurate brain tumor classification. While some of the existing solutions demonstrate noteworthy accuracy, the models become infeasible to deploy in areas where computational resources are limited. This research addresses the need for accurate and fast classification of brain tumors with a priority of deploying the model in technologically underdeveloped regions. The research presents a novel architecture for precise brain tumor classification fusing pretrained ResNet152V2 and modified VGG16 models. The proposed architecture undergoes a diligent fine-tuning process that ensures fine gradients are preserved in deep neural networks, which are essential for effective brain tumor classification. The proposed solution incorporates various image processing techniques to improve image quality and achieves an astounding accuracy of 98.36% and 98.04% in Figshare and Kaggle datasets respectively. This architecture stands out for having a streamlined profile, with only 2.8 million trainable parameters. We have leveraged 8-bit quantization to produce a model of size 73.881 MB, significantly reducing it from the previous size of 289.45 MB, ensuring smooth deployment in edge devices even in resource-constrained areas. Additionally, the use of Grad-CAM improves the interpretability of the model, offering insightful information regarding its decision-making process. Owing to its high discriminative ability, this model can be a reliable option for accurate brain tumor classification.

Figures (9)

• Figure 1: A sample of the Figshare and Kaggle dataset.
• Figure 2: The four-stage data preprocessing technique employed in the system.
• Figure 3: The proposed fusion architecture for brain tumor classification.
• Figure 4: Modified VGG16 network used in the fusion model.
• Figure 5: Confusion matrix of the fusion model with MLP head on different datasets.