A Novel DenseNet201-Based Deep Learning Framework for Multi-Class Brain Tumor Classification from MRI Scans

View at publisher Summarise with VIRAT
Authors: Hrithik Verma, Richa Sharma, Chamkaur Singh
Cited by: 0
References: 10

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Based on the DenseNet201 deep learning model, this study proposes a robust framework for early brain tumor detection, contributing to SDG 3 (Good Health and Well-Being) and SDG 9 (Industry, Innovation, and Infrastructure). Brain tumors, both benign and malignant, often result in severe neurological complications due to elevated intracranial pressure, making accurate and timely diagnosis essential. The research utilizes a dataset of 7,023 MRI scans categorized into pituitary tumors, gliomas, meningiomas, and no tumor. To improve model reliability and generalization, ImageDataGenerator was applied for data augmentation during the training and validation phases. The DenseNet201 architecture achieved notable performance in multi-class brain tumor classification, reaching an accuracy of 94.25% along with high precision, recall, and F1-scores. By enabling rapid AI-assisted diagnosis of brain tumors, this study enhances clinical decision-making, supports early intervention, and strengthens healthcare infrastructure through advanced, technology-driven medical solutions.

Fields of Study & Subjects

Brain Tumor Detection and Classification AI in cancer detection Glioma Diagnosis and Treatment

Publication details

Journal / source
International Conference on Signal Processing and Electronics Design (ICSPED)
Pages
121–125
Publication date
23 January 2026
Language
EN