Attention-Gated UNet Model for Semantic Segmentation of Brain Tumors and Feature Extraction for Survival Prognosis

Rut Patel, Snehal Rajput, Mehul S. Raval, Rupal A. Kapdi, Mohendra Roy · 2026

Read Attention-Gated UNet Model for Semantic Segmentation of Brain Tumors and Feature Extraction for Survival Prognosis by Rut Patel, Snehal Rajput, Mehul S. Raval, Rupal A. Kapdi, Mohendra Roy (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-15067.txt.

Description

Pate and colleagues build an attention-gated recurrent residual U-Net model for brain tumor segmentation and survival prognosis. Gliomas rank among the most common primary brain tumors and vary widely in aggression, outlook, and histology, complicating treatment and surgery. Their 2.5D model joins residual, recurrent, and three-plane designs to sharpen feature representation and segmentation quality while staying fast.

On the BraTS2021 validation set, whole tumor outlining reaches a Dice score of nine tenths, comparable to leading models. The network also extracts 64 features per plane model for outlook calls, cut to 28 with an artificial neural network. That outlook model reaches about 46 percent accuracy, a mean squared error near 108 thousand, and a Spearman rank correlation near one third on the test set.

Catalog record

CREATORRut Patel, Snehal Rajput, Mehul S. Raval, Rupal A. Kapdi, Mohendra Roy
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE32 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2602-15067

Browse the full library · About the Tower of Babel