An Automatic Identification of Brain Tumors in MRI Using Transfer Learning Approach

DOI

10.2991/978-94-6463-662-8_26How to use a DOI?

Keywords

magnetic resonance imaging; glioma; meningioma; pituitary gland; Inception V3

Abstract

Brain tumors are seen as formidable diseases that drastically reduce life expectancy. Patients who receive misdiagnoses or inadequate treatment for these tumors have a lower likelihood of survival. Magnetic resonance imaging (MRI) is often utilized for the analysis of tumors. However, the large volume of data generated by MRI makes manual segmentation within a reasonable timeframe challenging, thereby restricting the application of standard criteria in clinical settings. Hence, the need for efficient and automated segmentation techniques is evident. Detecting and accurately segmenting brain tumors early in biomedical imaging poses a significant challenge. The variability in temporal and anatomical features of brain tumors complicates automated segmentation. Early identification and treatment are vital for tackling this issue. A range of traditional machine learning (ML) algorithms has been utilized for the detection of brain tumors. Nevertheless, the dependence on features extracted manually serves as a significant drawback for these models.

The objective of this paper is to facilitate the early diagnosis of brain tumors through multi classification, employing convolutional neural networks (CNNs). The study presents three distinct CNN models tailored for three specific classification tasks related to brain tumor detection. The proposed CNN architecture was developed using a transfer learning model called Inception V3 can classify the brain tumor into three brain tumor types as glioma, meningioma and pituitary with training accuracy of 100% and validation accuracy of 97%.

The proposed CNN models are benchmarked against several widely recognized state-of-the-art CNN architectures including AlexNet, ResNet-50, VGG-16, and GoogleNet. Through experimentation with large, publicly available clinical datasets, the study demonstrates satisfactory classification results. The suggested CNN models serve as a helpful resource for doctors and radiologists in confirming their preliminary assessments of brain tumor multiclassification. By utilizing these models, healthcare professionals can improve the accuracy and efficiency of their diagnostic procedures, ultimately enhancing patient care outcomes.

Copyright

© 2025 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  - O. Roopa Devi
AU  - A. Vishnuvardhan Reddy
AU  - S. Shashikala
AU  - M. Mahesh Kumar
PY  - 2025
DA  - 2025/03/17
TI  - An Automatic Identification of Brain Tumors in MRI Using Transfer Learning Approach
BT  - Proceedings of the International Conference on Advanced Materials, Manufacturing and Sustainable Development (ICAMMSD 2024)
PB  - Atlantis Press
SP  - 319
EP  - 328
SN  - 2352-5401
UR  - https://doi.org/10.2991/978-94-6463-662-8_26
DO  - 10.2991/978-94-6463-662-8_26
ID  - Devi2025
ER  -