Machine Learning and Optical Coherence Tomography-Derived Radiomics Analysis to Predict Persistent Diabetic Macular Edema in Patients Undergoing Anti-Vascular Endothelial Growth Factor Intravitreal Therapy

Background: Diabetic macular edema (DME) is a leading cause of vision loss in patients with diabetes. Accurate predictions of individual treatment responses are needed. Here, we developed and evaluated an OCT-omics prediction model to assess anti-vascular endothelial growth factor (VEGF) treatment response in patients with DME. Methods: A retrospective analysis of 113 eyes from 82 patients with DME was conducted. Clinical and optical coherence tomography (OCT) data were collected; comprehensive feature engineering was performed. Logistic regression, support vector machine (SVM), and backpropagation neural network (BPNN) classifiers were trained using a training set of 79 eyes, and evaluated on a test set of 34 eyes. Clinical implications of the OCT-omics prediction model were assessed by decision curve analysis. Performance metrics were calculated, including sensitivity, specificity, F1 score, and area under the curve (AUC). Findings: The logistic, SVM, and BPNN classifiers demonstrated robust discriminative abilities in both the training and test sets. In the training set, the logistic classifier achieved a sensitivity of 0.904, specificity of 0.741, F1 score of 0.887, and AUC of 0.910. The SVM classifier showed a sensitivity of 0.923, specificity of 0.667, F1 score of 0.881, and AUC of 0.897. The BPNN classifier exhibited a sensitivity of 0.962, specificity of 0.926, F1 score of 0.962, and AUC of 0.982. Similar discriminative capabilities were maintained in the test set. OCT-omics scores were significantly higher in the non-persistent DME group than in the persistent DME group (p<0.001). OCT-omics scores were also positively correlated with the rate of decline in central subfield thickness after treatment (Pearson's R=0.44, p<0.001). Interpretation: Our OCT-omics prediction model demonstrated accurate assessment of anti-VEGF treatment response in patients with DME. It can provide valuable insights into individual patient prognoses and treatment outcomes. Our model's robust performance and potential clinical implications highlight its utility as a non-invasive tool for personalised treatment prediction and retinal pathology assessment. Funding: This project was funded by the Hunan Scientific Research Program of traditional Chinese Medicine (C2022009). Declaration of Interest: All authors declare no conflicts of interest. Ethical Approval: Ethical approval for this retrospective cohort study was obtained from the Ethics Committee of the Second Xiangya Hospital, Central South University (approval no. 2017-053). The study adhered to the principles established in the Declaration of Helsinki and other applicable ethical guidelines. To ensure privacy and comply with data protection regulations, patient data were anonymised and kept confidential.