Kurniasari, Dian and Maulana, Muhtarom Ahkam and Lumbanraja, Favorisen R and Warsono, Warsono (2025) Leveraging Transfer Learning Using Convolutional Neural Networks to Enhance Brain Tumor Images Classification. IAENG International Journal of Computer Science, 52 (11). pp. 4150-4163. ISSN 1819-9224

Text Leveraging Transfer Learning Using Convolutional Neural Networks to Enhance Brain Tumor Images Classification.pdf Download (1MB)

Abstract

Effective treatment relies on accurately diagnosing brain tumors, which are characterized by abnormal cell proliferation. Artificial Intelligence (AI) offers a promising alternative to traditional diagnostic methods, which are frequently error-prone. This study aims to enhance the precision of brain tumor image classification using a Transfer Learning (TL) approach with Convolutional Neural Networks (CNNs). A dataset of 7,020 images were categorized into four categories: glioma, meningioma, pituitary tumor, and no tumor. This dataset was used to test several pre-trained models, including DenseNet121, InceptionResNetV2, MobileNetV2, NasNetMobile, and ResNet50V2. Performance was measured using accuracy, precision, sensitivity, and specificity metrics. The most effective of these was ResNet50V2, which achieved an accuracy of 97.70% and a loss of 0.066. A confusion matrix analysis of the results highlighted the model's exceptional performance, with sensitivity (97.70%), specificity (99.30%), and precision (97.80%). This research significantly contributes to medical image analysis, improving diagnostic accuracy using AI technology. The application of TL enhances early detection reduces and reduces the misdiagnosis by lowering the need for large datasets and minimizing errors. Furthermore, the model's efficiency in analyzing large numbers of MRI images significantly offers time-saving advantages for healthcare professionals, allowing them to prioritize more complex cases. This study advances the role of AI in medical diagnostics, particularly in brain tumor classification, with the potential to revolutionize early diagnosis, treatment strategies, and expand access to quality healthcare in underserved areas. By improving diagnostic accuracy, this model could contribute to reducing treatment delays, ultimately saving more lives.

Item Type:	Article
Subjects:	Q Science > QA Mathematics
Divisions:	Fakultas Matematika dan Ilmu Pengetahuan Alam (FMIPA) > Prodi Matematika
Depositing User:	DIAN KURNIASARI
Date Deposited:	17 Apr 2026 02:42
Last Modified:	17 Apr 2026 02:42
URI:	http://repository.lppm.unila.ac.id/id/eprint/54827

Actions (login required)

View Item