Whole-orbit radiomics: machine learning-based multi- and fused- region radiomics signatures for intravenous glucocorticoid response prediction in thyroid eye disease

• Published in: Journal of translational medicine Vol. 22; no. 1; pp. 56 - 14
• Main Authors: Zhang, Haiyang, Jiang, Mengda, Chan, Hoi Chi, Zhang, Huijie, Xu, Jiashuo, Liu, Yuting, Zhu, Ling, Tao, Xiaofeng, Xia, Duojin, Zhou, Lei, Li, Yinwei, Sun, Jing, Song, Xuefei, Zhou, Huifang, Fan, Xianqun
• Format: Journal Article
• Language: English
• Published: London BioMed Central 13.01.2024; BioMed Central Ltd; BMC
• Subjects: Algorithms; Biomedical and Life Sciences; Biomedicine; Care and treatment; Corticosteroids; Diagnosis; Diplopia; Dosage and administration; Edema; Eye diseases; Glucocorticoids; Glucocorticoids - therapeutic use; Graves Ophthalmopathy - diagnostic imaging; Health aspects; Humans; Image processing; Intravenous administration; Intravenous glucocorticoid; Lacrimal gland and Nasolacrimal duct; Learning algorithms; Machine Learning; Magnetic resonance imaging; Magnetic Resonance Imaging - methods; Medicine/Public Health; Methods; MRI; Multi-organ segmentation; Muscles; Oculomotor system; Optic nerve; Orbit - diagnostic imaging; Pathogenesis; Prediction models; Radiomics; Radiomics analysis; Response prediction; Retrospective Studies; Segmentation; Soft tissues; Thyroid; Thyroid eye disease; Thyroid gland; Translational Ophthalmology; Visual acuity; China; Radiomics analysis; Thyroid eye disease; MRI; Multi-organ segmentation; Response prediction; Intravenous glucocorticoid
• ISSN: 1479-5876; 1479-5876
• DOI: 10.1186/s12967-023-04792-2

Background Radiomics analysis of orbital magnetic resonance imaging (MRI) shows preliminary potential for intravenous glucocorticoid (IVGC) response prediction of thyroid eye disease (TED). The current region of interest segmentation contains only a single organ as extraocular muscles (EOMs). It would be of great value to consider all orbital soft tissues and construct a better prediction model. Methods In this retrospective study, we enrolled 127 patients with TED that received 4·5 g IVGC therapy and had complete follow-up examinations. Pre-treatment orbital T2-weighted imaging (T2WI) was acquired for all subjects. Using multi-organ segmentation (MOS) strategy, we contoured the EOMs, lacrimal gland (LG), orbital fat (OF), and optic nerve (ON), respectively. By fused-organ segmentation (FOS), we contoured the aforementioned structures as a cohesive unit. Whole-orbit radiomics (WOR) models consisting of a multi-regional radiomics (MRR) model and a fused-regional radiomics (FRR) model were further constructed using six machine learning (ML) algorithms. Results The support vector machine (SVM) classifier had the best performance on the MRR model (AUC = 0·961). The MRR model outperformed the single-regional radiomics (SRR) models (highest AUC = 0·766, XGBoost on EOMs, or LR on OF) and conventional semiquantitative imaging model (highest AUC = 0·760, NaiveBayes). The application of different ML algorithms for the comparison between the MRR model and the FRR model (highest AUC = 0·916, LR) led to different conclusions. Conclusions The WOR models achieved a satisfactory result in IVGC response prediction of TED. It would be beneficial to include more orbital structures and implement ML algorithms while constructing radiomics models. The selection of separate or overall segmentation of orbital soft tissues has not yet attained its final optimal result.