Comparative Analysis of Resource-Efficient CNN Architectures for Brain Tumor Classification

TL;DR

The paper addresses the challenge of accurate brain tumor classification in MRI while reducing computational burden. It introduces two lightweight CNNs, BTBCNN and BTMCNN, and benchmarks them against ResNet-18 and VGG-16 on Br35H and Brain Tumor MRI Dataset, including full-finetuning and few-shot learning. Results show that the custom CNNs achieve competitive accuracy with far fewer parameters and faster inference, while pretrained models retain a slight edge in some cases. The work demonstrates the practicality of resource-efficient architectures for medical imaging, especially in resource-limited clinical settings, and encourages further exploration of lightweight designs and data-efficient learning.

Abstract

Accurate brain tumor classification in MRI images is critical for timely diagnosis and treatment planning. While deep learning models like ResNet-18, VGG-16 have shown high accuracy, they often come with increased complexity and computational demands. This study presents a comparative analysis of effective yet simple Convolutional Neural Network (CNN) architecture and pre-trained ResNet18, and VGG16 model for brain tumor classification using two publicly available datasets: Br35H:: Brain Tumor Detection 2020 and Brain Tumor MRI Dataset. The custom CNN architecture, despite its lower complexity, demonstrates competitive performance with the pre-trained ResNet18 and VGG16 models. In binary classification tasks, the custom CNN achieved an accuracy of 98.67% on the Br35H dataset and 99.62% on the Brain Tumor MRI Dataset. For multi-class classification, the custom CNN, with a slight architectural modification, achieved an accuracy of 98.09%, on the Brain Tumor MRI Dataset. Comparatively, ResNet18 and VGG16 maintained high performance levels, but the custom CNNs provided a more computationally efficient alternative. Additionally,the custom CNNs were evaluated using few-shot learning (0, 5, 10, 15, 20, 40, and 80 shots) to assess their robustness, achieving notable accuracy improvements with increased shots. This study highlights the potential of well-designed, less complex CNN architectures as effective and computationally efficient alternatives to deeper, pre-trained models for medical imaging tasks, including brain tumor classification. This study underscores the potential of custom CNNs in medical imaging tasks and encourages further exploration in this direction.

Figures (6)

• Figure 1: Brain Tumor MRI sample images.
• Figure 2: Overview of the stages in the methodology: data preparation, model building, training and evaluation, and few-shot learning.
• Figure 3: A summary of the training and testing data distribution for the Br35H and Brain Tumor MRI Dataset, along with the class distribution of MRI images within the Brain Tumor MRI Dataset.
• Figure 4: Confusion matrices for binary classification using BTBCNN on the Br35H and Brain Tumor MRI Dataset, and for multi-class classification using BTMCNN on the Brain Tumor MRI Dataset of the Full-Finetuning Experiment.
• Figure 5: Performance comparison of custom CNNs (BTBCNN, BTMCNN) and pre-trained models (ResNet18, VGG16) in terms of accuracy, precision, recall, and F1-scores for binary and multi-class classification tasks.