Proceedings of the 1st Engineering Data Analytics and Management Conference (EAMCON 2025)

<Previous Article In Volume
Next Article In Volume>

Hybrid Agentic Vision Transformer with Hummingbird-Optimized Patch-wise M-Net for Advanced Brain Tissue Segmentation in MRI

Authors
Srirangam Bhavani1, 2, *, N. Subhash Chandra1
1Dept. of CSE, CVR College of Engineering, Hyderabad, India
2Dept. of CSE, JNTU, Hyderabad, India
*Corresponding author. Email: s.bhavani@cvr.ac.in
Corresponding Author
Srirangam Bhavani
Available Online 31 December 2025.
DOI
10.2991/978-94-6463-978-0_22How to use a DOI?
Keywords
Brain tissue segmentation; MRI; Vision Transformer; Agentic AI; M-Net; Hummingbird optimization; Deep learning
Abstract

Accurate segmentation of brain tissues (White Matter, Gray Matter, and Cerebrospinal Fluid) is critical for neurological diagnosis. This paper presents a novel Hybrid Agentic Vision Trans-former Framework with Hummingbird-Optimized Patchwise M-Net (HAVT-HOM) that synergistically combines: (1) patch-wise M-Net optimized via Hummingbird Optimization Algorithm for local feature extraction, (2) Vision Transformer for global contextual modeling, and (3) an Agentic AI layer (i.e., an AI system guided by intelligent agents that can reason and use external knowledge) incorporating Retrieval-Augmented Generation and Context-Augmented Generation for adaptive refinement. We introduce the Multi-Scale Fusion Module, Agent-Modulated Cross-Attention mechanism, and Context-Driven Iterative Refinement Protocol. Rigorous 10-fold cross-validation on the IBSR dataset demonstrates HAVT-HOM achieves 97.85 0.18% Dice coefficient, representing a 1.40% improvement over baseline, with superior robustness to imaging artifacts and boundary precision.

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)

<Previous Article In Volume
Next Article In Volume>
Volume Title
Proceedings of the 1st Engineering Data Analytics and Management Conference (EAMCON 2025)
Series
Advances in Engineering Research
Publication Date
31 December 2025
ISBN
978-94-6463-978-0
ISSN
2352-5401
DOI
10.2991/978-94-6463-978-0_22How to use a DOI?
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.

Cite this article

risenwbib
TY  - CONF
AU  - Srirangam Bhavani
AU  - N. Subhash Chandra
PY  - 2025
DA  - 2025/12/31
TI  - Hybrid Agentic Vision Transformer with Hummingbird-Optimized Patch-wise M-Net for Advanced Brain Tissue Segmentation in MRI
BT  - Proceedings of the 1st Engineering Data Analytics and Management Conference (EAMCON 2025)
PB  - Atlantis Press
SP  - 239
EP  - 249
SN  - 2352-5401
UR  - https://doi.org/10.2991/978-94-6463-978-0_22
DO  - 10.2991/978-94-6463-978-0_22
ID  - Bhavani2025
ER  -