Gaze-Driven Intelligence for Brain Tumor Detection
International Journal of Advanced Research in Science and Technology · 2026 · 15 (5) · 1928-1932
DOI: 10.62226/ijarst20262671 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Advanced Research in Science and Technology, 2026.
Abstract (Originalsprache)
Brain tumors continue to be one of the deadliest neurological conditions, and the core problem is not just the disease itself but how late it tends to get caught. Patients usually walk into a clinic after symptoms have already become severe, by which point the tumor may have progressed significantly. This paper describes a two-phase AI-based diagnostic system that uses gaze data captured through a normal webcam as a preliminary screening tool, followed by MRI analysis using a Patch-Based Vision Transformer (PBViT) for tumor classification. Phase 1 employs a CNN to analyze ocular biomarkers pupil dilation, blink rate, and inter-eye asymmetry. Phase 2 uses PBViT on uploaded brain scans to classify tumor type and grade. The proposed system achieved an overall detection accuracy of 94.6%, offering a practical and low-cost pathway for early neurological screening.
Zitieren
APA
(2026). Gaze-Driven Intelligence for Brain Tumor Detection. International Journal of Advanced Research in Science and Technology, 15(5), 1928-1932. https://doi.org/10.62226/ijarst20262671
Vancouver
Gaze-Driven Intelligence for Brain Tumor Detection. International Journal of Advanced Research in Science and Technology. 2026;15(5):1928-1932. doi:10.62226/ijarst20262671
BibTeX
@article{euvimed2026,
title = {Gaze-Driven Intelligence for Brain Tumor Detection},
journal = {International Journal of Advanced Research in Science and Technology},
year = {2026},
volume = {15},
number = {5},
pages = {1928--1932},
doi = {10.62226/ijarst20262671}
}
EUVIMED