Application of a multimodal model integrating optical coherence tomography angiography (OCTA) and radiomics for the identification of high-risk Nonarteritic anterior ischaemic optic neuropathy (NAION) populations and for distinguishing its acute phase.
5 affiliations
- 1First Affiliated Hospital of Kunming Medical University, Department of Ophthalmology, Kunming, China.
- 2Department of Ophthalmology, Affiliated Hospital of North Sichuan Medical College, Nanchong, China.
- 3School of Information Engineering and Automation, Kunming University of Science and Technology, Kunming, China.
- 4School of Information Engineering and Automation, Kunming University of Science and Technology, Kunming, China. [email protected].
- 5First Affiliated Hospital of Kunming Medical University, Department of Ophthalmology, Kunming, China. [email protected].
Abstract
Background
Nonarteritic anterior ischaemic optic neuropathy (NAION) is an acute optic neuropathy characterized primarily by sudden vision loss accompanied by visual field defects. Early screening and diagnosis are especially critical because of the diagnostic challenges associated with the initial stages of NAION and its irreversible nature. This study proposes a multimodal diagnostic approach that integrates optical coherence tomography angiography (OCTA) and radiomics for the identification of high-risk NAION populations and for distinguishing its acute phase.Methods
Without requiring region-of-interest (ROI) delineation, we globally extracted radiomics features from OCTA images and combined them with clinical information. Multilayer perceptron (MLP) models were subsequently developed for three distinct scenarios: single-modality single-region, multimodality single-region, and multimodality multiregion analyses.Results
The results demonstrate that the multimodality multiregion model, incorporating multiregional OCTA features and clinical information, achieved optimal performance in NAION prediction, with an AUC of 0.996 (95% CI, 0.983-1.000) and an ACC of 0.970. This model also exhibited outstanding performance in differentiating acute-phase NAION from acute-phase optic neuritis (ON), with an AUC of 0.989 (95% CI, 0.966-1.000) and an ACC of 0.944.Conclusions
By integrating multisource information, the model not only enhances diagnostic accuracy but also provides clinicians with comprehensive decision support perspectives. This advancement facilitates early intervention and personalized treatment strategies, thereby mitigating disease progression and improving prognostic outcomes for NAION patients. This study establishes a novel diagnostic pathway and offers innovative methodologies for the early diagnosis and management of NAION.
Related
- Nomogram models integrating multiregion radiomics, quantitative computed tomography, and clinical indicators for distinguishing connective tissue disease-associated interstitial lung disease from idiopathic pulmonary fibrosis and improving the prediction of gender-age-physiology stage
- Comprehensive coronary computed tomography angiography assessment in high-altitude diabetic populations: integrating plaque characteristics, pericoronary inflammation, and coronary function
- Perivascular fat attenuation index improves the identification of functional myocardial ischemia beyond traditional risk factors: a coronary CT angiography-derived fractional flow reserve study
- Comprehensive identification of lactate metabolism-related genes in recurrent spontaneous abortion : An observational study integrating bioinformatics and experimental validation
- The application of explainable machine learning for the identification of MASLD in non-obese populations
BibTeX
@article{yang2026application,
title = {{Application of a multimodal model integrating optical coherence tomography angiography (OCTA) and radiomics for the identification of high-risk Nonarteritic anterior ischaemic optic neuropathy (NAION) populations and for distinguishing its acute phase.}},
author = {Peixian Yang and Songlin Cui and Yuxin Zhang and Ruijuan Zhao and Xiaoting Xi and Xin Xiong and Luda Chen and Jianfeng He and Jia Ma},
journal = {BMC ophthalmology},
year = {2026},
month = aug,
doi = {10.1186/s12886-026-05186-w},
url = {https://europepmc.org/article/MED/42958118},
note = {CC BY},
}
RIS
TY - JOUR
TI - Application of a multimodal model integrating optical coherence tomography angiography (OCTA) and radiomics for the identification of high-risk Nonarteritic anterior ischaemic optic neuropathy (NAION) populations and for distinguishing its acute phase.
AU - Peixian Yang
AU - Songlin Cui
AU - Yuxin Zhang
AU - Ruijuan Zhao
AU - Xiaoting Xi
AU - Xin Xiong
AU - Luda Chen
AU - Jianfeng He
AU - Jia Ma
PY - 2026
DA - 2026/08/03
JO - BMC ophthalmology
DO - 10.1186/s12886-026-05186-w
UR - https://europepmc.org/article/MED/42958118
AB - Background Nonarteritic anterior ischaemic optic neuropathy (NAION) is an acute optic neuropathy characterized primarily by sudden vision loss accompanied by visual field defects. Early screening and diagnosis are especially critical because of the diagnostic challenges associated with the initial stages of NAION and its irreversible nature. This study proposes a multimodal diagnostic approach that integrates optical coherence tomography angiography (OCTA) and radiomics for the identification of high-risk NAION populations and for distinguishing its acute phase.Methods Without requiring region-of-interest (ROI) delineation, we globally extracted radiomics features from OCTA images and combined them with clinical information. Multilayer perceptron (MLP) models were subsequently developed for three distinct scenarios: single-modality single-region, multimodality single-region, and multimodality multiregion analyses.Results The results demonstrate that the multimodality multiregion model, incorporating multiregional OCTA features and clinical information, achieved optimal performance in NAION prediction, with an AUC of 0.996 (95% CI, 0.983-1.000) and an ACC of 0.970. This model also exhibited outstanding performance in differentiating acute-phase NAION from acute-phase optic neuritis (ON), with an AUC of 0.989 (95% CI, 0.966-1.000) and an ACC of 0.944.Conclusions By integrating multisource information, the model not only enhances diagnostic accuracy but also provides clinicians with comprehensive decision support perspectives. This advancement facilitates early intervention and personalized treatment strategies, thereby mitigating disease progression and improving prognostic outcomes for NAION patients. This study establishes a novel diagnostic pathway and offers innovative methodologies for the early diagnosis and management of NAION.
SN - 1471-2415
ER -