Clinically-aligned ischemic stroke segmentation and ASPECTS scoring on NCCT imaging using a slice-gated loss on foundation representations

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Azeem, Hiba, Khan, Behraj, Syed, Tahir Qasim
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914357679489024
author Azeem, Hiba
Khan, Behraj
Syed, Tahir Qasim
author_facet Azeem, Hiba
Khan, Behraj
Syed, Tahir Qasim
contents Rapid infarct assessment on non-contrast CT (NCCT) is essential for acute ischemic stroke management. Most deep learning methods perform pixel-wise segmentation without modeling the structured anatomical reasoning underlying ASPECTS scoring, where basal ganglia (BG) and supraganglionic (SG) levels are clinically interpreted in a coupled manner. We propose a clinically aligned framework that combines a frozen DINOv3 backbone with a lightweight decoder and introduce a Territory-Aware Gated Loss (TAGL) to enforce BG-SG consistency during training. This anatomically informed supervision adds no inference-time complexity. Our method achieves a Dice score of 0.6385 on AISD, outperforming prior CNN and foundation-model baselines. On a proprietary ASPECTS dataset, TAGL improves mean Dice from 0.698 to 0.767. These results demonstrate that integrating foundation representations with structured clinical priors improves NCCT stroke segmentation and ASPECTS delineation.
format Preprint
id arxiv_https___arxiv_org_abs_2602_23961
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Clinically-aligned ischemic stroke segmentation and ASPECTS scoring on NCCT imaging using a slice-gated loss on foundation representations
Azeem, Hiba
Khan, Behraj
Syed, Tahir Qasim
Image and Video Processing
Computer Vision and Pattern Recognition
Rapid infarct assessment on non-contrast CT (NCCT) is essential for acute ischemic stroke management. Most deep learning methods perform pixel-wise segmentation without modeling the structured anatomical reasoning underlying ASPECTS scoring, where basal ganglia (BG) and supraganglionic (SG) levels are clinically interpreted in a coupled manner. We propose a clinically aligned framework that combines a frozen DINOv3 backbone with a lightweight decoder and introduce a Territory-Aware Gated Loss (TAGL) to enforce BG-SG consistency during training. This anatomically informed supervision adds no inference-time complexity. Our method achieves a Dice score of 0.6385 on AISD, outperforming prior CNN and foundation-model baselines. On a proprietary ASPECTS dataset, TAGL improves mean Dice from 0.698 to 0.767. These results demonstrate that integrating foundation representations with structured clinical priors improves NCCT stroke segmentation and ASPECTS delineation.
title Clinically-aligned ischemic stroke segmentation and ASPECTS scoring on NCCT imaging using a slice-gated loss on foundation representations
topic Image and Video Processing
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2602.23961