Deep Learning-Based Multi-Class Brain Tumor Classification from MRI Images Using DenseNet121 & OpenCV
Journal of Sensors, IoT & Health Sciences · 30.06.2026 · 4 (2) · 9-17
DOI: 10.69996/jsihs.2026007 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Sensors, IoT & Health Sciences, online erschienen am 30.06.2026. Verfasst von Jasti Kumari und 1 weiteren Autor:innen.
Abstract (Originalsprache)
Brain tumors are among the most life-threatening neurological conditions. Early detection through medical imaging plays a critical role in improving patient outcomes, but manual analysis of MRI scans is time-intensive and subject to human error. This project addressesthat problem by building a desktop application that can automatically classify brain MRI images into four categories: Glioma, Meningioma, Pituitary tumor, and No Tumor. The classification model is based on DenseNet121, a deep convolutional neural network architecture known for its dense connectivity pattern that improves gradient flow and feature reuse across layers. The model was trained on a labeled dataset containing thousands of MRI images and achieved an overall accuracy of 83% on a held-out test set of more than 1,300 images. Among the individual classes, No Tumor detection achieves the highest accuracy at 93%, followed by Pituitary at 88%, Glioma at 83%, and Meningioma at 65%. Once trained, the Keras model was exported to the ONNX format, which allows it to run through OpenCV's DNN module without requiring Tensor Flow or any GPU-specific libraries at runtime. This makes the application significantly lighter and more portable than typical deep learning deployments.
Zitieren
APA
Kumari, J., & Navya, P. S. (2026). Deep Learning-Based Multi-Class Brain Tumor Classification from MRI Images Using DenseNet121 & OpenCV. Journal of Sensors, IoT & Health Sciences, 4(2), 9-17. https://doi.org/10.69996/jsihs.2026007
Vancouver
Kumari J, Navya PS. Deep Learning-Based Multi-Class Brain Tumor Classification from MRI Images Using DenseNet121 & OpenCV. Journal of Sensors, IoT & Health Sciences. 2026;4(2):9-17. doi:10.69996/jsihs.2026007
BibTeX
@article{Kumari2026,
author = {Kumari, Jasti and Navya, Parchuri Swetha},
title = {Deep Learning-Based Multi-Class Brain Tumor Classification from MRI Images Using DenseNet121 & OpenCV},
journal = {Journal of Sensors, IoT & Health Sciences},
year = {2026},
volume = {4},
number = {2},
pages = {9--17},
doi = {10.69996/jsihs.2026007}
}
EUVIMED