TY - JOUR
T1 - Multi-Parametric Optical Coherence Tomography Angiography Quantifies Heterogeneous Therapeutic Responses in Port Wine Stains
AU - Zhang, Xiwan
AU - Chen, Defu
AU - Li, Jinbin
AU - Yang, Xiaoyu
AU - Liu, Yidi
AU - Qiu, Haixia
AU - Gu, Ying
N1 - Publisher Copyright:
© 1995-2012 IEEE.
PY - 2026
Y1 - 2026
N2 - Port wine stains (PWS), congenital vascular malformation, affecting 3 ‰-5 ‰ of newborns, pose significant psychological and social challenges. Current treatments, including vascular-targeted photodynamic therapy (V-PDT), exhibit variable efficacy due to interpatient heterogeneity in vascular characteristics. Here, we present a multi-parametric quantitative method using optical coherence tomography angiography (OCTA) to characterize PWS lesions, analyzing 26 vascular indicators in 100 patients. Our method effectively discriminates patients with distinct therapeutic responses and identifies morphological vascular changes linked to treatment outcomes. By integrating quantitative vascular data with clinical parameters via machine learning, we developed a predictive model that distinguished no-response (NR) from moderate-improvement (MI) and high-improvement (HI) groups with accuracies of 75% and 91%, respectively. This study highlights the importance of understanding vascular pathology to advance personalized treatment strategies for PWS, offering a novel framework for non-invasive evaluation and management of vascular lesions.
AB - Port wine stains (PWS), congenital vascular malformation, affecting 3 ‰-5 ‰ of newborns, pose significant psychological and social challenges. Current treatments, including vascular-targeted photodynamic therapy (V-PDT), exhibit variable efficacy due to interpatient heterogeneity in vascular characteristics. Here, we present a multi-parametric quantitative method using optical coherence tomography angiography (OCTA) to characterize PWS lesions, analyzing 26 vascular indicators in 100 patients. Our method effectively discriminates patients with distinct therapeutic responses and identifies morphological vascular changes linked to treatment outcomes. By integrating quantitative vascular data with clinical parameters via machine learning, we developed a predictive model that distinguished no-response (NR) from moderate-improvement (MI) and high-improvement (HI) groups with accuracies of 75% and 91%, respectively. This study highlights the importance of understanding vascular pathology to advance personalized treatment strategies for PWS, offering a novel framework for non-invasive evaluation and management of vascular lesions.
KW - Optical coherence tomography angiography (OCTA)
KW - port wine stains
KW - therapeutic effect prediction
KW - vascular-targeted photodynamic therapy (V-PDT)
UR - https://www.scopus.com/pages/publications/105015957227
U2 - 10.1109/JSTQE.2025.3609623
DO - 10.1109/JSTQE.2025.3609623
M3 - Article
AN - SCOPUS:105015957227
SN - 1077-260X
VL - 32
JO - IEEE Journal of Selected Topics in Quantum Electronics
JF - IEEE Journal of Selected Topics in Quantum Electronics
IS - 4
M1 - 7200214
ER -