Abstract
Radiation-induced lung injury (RILI) is a prevalent dose-limiting toxicity in thoracic radiotherapy. Current risk assessment strategies primarily rely on do-simetric variables and imaging biomarkers, while the vascular architecture bi-omarkers derived from the pulmonary vasculature and tumor–vessel spatial in-teractions remain insufficiently explored. Given the critical role of vascular in-jury in the pathogenesis of RILI, this study aimed to investigate potential im-aging biomarkers at both the pulmonary vascular scale and the tumor–vessel local scale. A retrospective cohort consisting of 154 patients who received tho-racic radiotherapy was enrolled, including 123 cases with early RILI and 31 cases without RILI. Leveraged pre-treatment non-contrast average CT, pulmo-nary vessel masks were automatically segmented by pretrained model and sub-sequently reconstructed into centerline networks and 3D surface models, from which a panel of 15 candidate structural biomarkers was extracted. Compara-tive analysis via the Mann–Whitney U test identified the vascular fractal di-mension (FD) as the strongest biomarker. Patients with RILI exhibited signifi-cantly lower FD values compared to controls (2.1925 vs. 2.2691, P = 0.0055). Notably, FD retained the highest statistical significance after FDR adjustment, underscoring its robust discriminatory capacity. This study demonstrates that decreased structural complexity of the pulmonary vasculature on routine non-contrast planning CT is closely associated with higher susceptibility to early RILI, indicating the clinical potential of this quantitative biomarker for early RILI risk stratification.
Links to Paper and Supplementary Materials
Main Paper (Open Access Version): https://papers.miccai.org/miccai-2026-sat/paper/CMMCA2026_014.pdf
SharedIt Link: Not yet available
SpringerLink (DOI): Not yet available
Supplementary Material: Not Submitted
Link to Open Review
Open Review Page: https://openreview.net/forum?id=xOPu4T1bHI
BibTex
@InProceedings{HuaWen_Pulmonary_MICCAISAT2026,
author = { Huang, Wenyuan AND Ma, Lixin AND He, Jiahui AND Jiang, Xin AND Qin, Wenjian AND Gu, Jia},
title = { { Pulmonary Vascular Fractal Dimension on Non-Contrast Planning CT as a Potential Imaging Biomarker for Early Radiation-Induced Lung Injury: A Multi-Scale Structural Analysis } },
booktitle = {Medical Image Computing and Computer Assisted Intervention -- MICCAI 2026 Workshops and Challenges},
year = {2026},
publisher = {Springer Nature Switzerland},
volume = {LNCS 17278},
month = {pending},
page = {pending}
}