Adaptive Inter-Region Attention Network for Multimodal Brain Tumor Analysis
International Journal of Engineering & Extended Technologies Research · 28.03.2026 · 08 (02)
DOI: 10.15662/ijeetr.2026.0802387 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Engineering & Extended Technologies Research, online erschienen am 28.03.2026. Verfasst von V. Indhumathi und 4 weiteren Autor:innen.
Abstract (Originalsprache)
Functional Magnetic Resonance Imaging (fMRI) provides critical insights into dynamic brain activity and connectivity patterns, enabling the study of neural functions over time. Traditional clustering methods, such as Topological Data Analysis (TDA), are limited to grouping brain networks and cannot accurately detect or classify active regions or abnormalities like tumors. Manual interpretation of MRI and fMRI data is time-consuming, prone to human error, and often inconsistent across multi-site datasets. To overcome these challenges, the proposed system introduces a hybrid YOLOv8 and Convolutional Neural Network (CNN) framework for automated detection, localization, classification, and staging of brain tumors along with analysis of normal brain activity. YOLOv8 precisely detects and localizes active brain regions and tumor regions in MRI and fMRI-derived maps, generating bounding boxes and confidence scores. The CNN extracts deep spatial and temporal features from these detected regions, enabling accurate classification of brain states and determination of tumor stage. The hybrid approach combines object detection with deep feature learning, capturing both spatial location and functional significance of regions. The system is trained on multi-site, multi-modal datasets to ensure robustness against scanner variability and non-neural artifacts. Experimental results show that the framework achieves high accuracy in both brain activity analysis and tumor detection and staging. This automated method reduces dependence on manual interpretation, providing faster, more reliable, and interpretable insights. It supports early diagnosis, treatment planning, and monitoring of brain tumors. By integrating YOLOv8 localization with CNN feature extraction, the system offers a comprehensive and scalable solution for brain network analysis and clinical applications. Overall, this approach represents a significant advancement in deep learning–based neuroimaging for automated brain activity and tumor assessment.
Zitieren
APA
Indhumathi, V., A, V., Kumar, V., C, T., & S, T. (2026). Adaptive Inter-Region Attention Network for Multimodal Brain Tumor Analysis. International Journal of Engineering & Extended Technologies Research, 08(02). https://doi.org/10.15662/ijeetr.2026.0802387
Vancouver
Indhumathi V, A V, Kumar V, C T, S T. Adaptive Inter-Region Attention Network for Multimodal Brain Tumor Analysis. International Journal of Engineering & Extended Technologies Research. 2026;08(02). doi:10.15662/ijeetr.2026.0802387
BibTeX
@article{Indhumathi2026,
author = {Indhumathi, V. and A, Vishnupriyan and Kumar, Vicky and C, Thirupathi and S, Theja},
title = {Adaptive Inter-Region Attention Network for Multimodal Brain Tumor Analysis},
journal = {International Journal of Engineering & Extended Technologies Research},
year = {2026},
volume = {08},
number = {02},
doi = {10.15662/ijeetr.2026.0802387}
}
EUVIMED