Enhancing Brain Tumor Classification with Explainable AI: A Comparative Study of CNN Model

DOI

10.2991/978-94-6239-791-0_9How to use a DOI?

Keywords

Medical Image Analysis; MRI; Brain Tumor; Deep Learning; CNN; ResNet50; EfficientNet; DenseNet; Inception V3; XAI

Abstract

Classification of brain tumor from MRI is among the foremost important processes within the healthcare industry because it facilitates early diagnosis of the patients, which further improves their quality of life. The objective is to investigate the application of deep learning methods using the CNN architecture along with the XAI for brain tumors classification in multi-class. In this regard, transfer learning technique has been adopted in this research work in order to employ four pre-trained deep learning models known as InceptionV3, EfficientNetB0, DenseNet121, and ResNet50. These models have been tested by using a balanced dataset of MRI images which included glioma, meningioma, pituitary tumor, and no tumor with a total number of 7,200 images. Feature extraction and fine-tuning were used for the training of all four models within the target domain. It was observed that ResNet50 yielded the most desirable outcomes among the four tested models with the highest training accuracy rate of 98.09% and test accuracy of 92.06% alongside precision, recall, and F1-score. To solve the problem of interpretability of deep learning models, Grad-CAM technique has been applied to the model having the best performances. This work represents an effort towards developing efficient and interpretable AI-powered diagnostic systems for the diagnosis of brain tumors.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Vishvajeet Choudhary
AU  - Yogita D. Patil
AU  - Karthik Kurup
AU  - Tanishka Gophane
AU  - Mahesh Kapse
AU  - Pushkar Kumar
PY  - 2026
DA  - 2026/10/09
TI  - Enhancing Brain Tumor Classification with Explainable AI: A Comparative Study of CNN Model
BT  - Proceedings of the International Conference on Artificial Intelligence and Machine Learning for Agro-Ecosystems (ICAIML-AE 2026)
PB  - Atlantis Press
SP  - 86
EP  - 95
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6239-791-0_9
DO  - 10.2991/978-94-6239-791-0_9
ID  - Choudhary2026
ER  -