Interpretable Machine Learning for Course Prediction in Central Serous Chorioretinopathy Using Baseline OCT from a Multicenter Cohort
American Journal of Ophthalmology
Abstract
PURPOSE: To predict disease clinical course in central serous chorioretinopathy (CSC) using an interpretable machine learning model applied to baseline artificial intelligence (AI)-derived optical coherence tomography (OCT) biomarkers acquired at first presentation. DESIGN: Multicenter, retrospective cohort study. PARTICIPANTS: Overall, 329 treatment-naïve CSC eyes recruited across eight international centers in Europe, the Middle East, Asia, and North Africa. METHODS: Treatment-naïve CSC patients were retrospectively included and classified, based on follow-up, as acute or chronic CSC according to subretinal fluid (SRF) resolution. Baseline OCT biomarkers were extracted using an AI-based platform (Discovery OCT Biomarker Detector; RetinAI AG, Switzerland), which quantified retinal layer thicknesses, including the outer nuclear layer (ONL) and photoreceptor-retinal pigment epithelium (PR+RPE) complex, fluid volumes (SRF and RPE detachment [PED]), and biomarker probabilities, across central, parafoveal, and perifoveal ETDRS regions. These features used to develop an interpretable machine learning model predicting future CSC chronicity. MAIN OUTCOME MEASURES: Classification of eyes as acute or chronic CSC, determined by SRF resolution on follow-up; discriminative performance of the Random Forest classifier, assessed by area under the receiver operating characteristic curve (AUC); and biomarker importance, evaluated using SHapley Additive exPlanations (SHAP) values. RESULTS: Of 329 eyes (291 patients), 92 (28.0%) were classified as acute and 237 (72.0%) as chronic CSC; mean follow-up was 4.2 ± 1.6 years. Chronic CSC patients were significantly older (51.7 ± 12.6 vs. 43.2 ± 9.6 years, p<0.001) with worse baseline visual acuity (0.25 ± 0.21 vs. 0.16 ± 0.15 logMAR, p=0.029). Among baseline OCT biomarkers, parafoveal ONL thickness showed the strongest association with disease chronicity (q<0.0001; r=-0.414). The Random Forest classifier achieved an AUC of 0.852 (95% CI: 0.748-0.940). SHAP analysis identified parafoveal ONL thickness, perifoveal PR+RPE thickness, and central SRF volume as the strongest predictors of acute CSC, whereas greater RNFL thickness, PED volume, FLAT PED probability, and HF probability were the most influential predictors of chronic disease. CONCLUSIONS: Automatically detected OCT biomarkers can predict CSC chronicity at initial presentation with good discriminatory performance. Quantitative markers of outer retinal integrity, fluid configuration, and RPE dysfunction may improve prognostication, facilitate patient counseling, and help identify individuals at increased risk of persistent CSC.
Authors 23
-
Affiliation as printed
Department of Ophthalmology, Inselspital, Bern University Hospital, University of Bern, Freiburgstrasse 15, CH-3010, Bern, Switzerland; Bern Photographic Reading Center, Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland. Electronic address: lorenzoferrodes@gmail.com
-
Affiliation as printed
Department of Ophthalmology, Inselspital, Bern University Hospital, University of Bern, Freiburgstrasse 15, CH-3010, Bern, Switzerland; ARTORG Research Center Biomedical Engineering Research, University of Bern, Bern, Switzerland
-
University of Bern · University Hospital of Bern
Affiliation as printed
Department of Ophthalmology, Inselspital, Bern University Hospital, University of Bern, Freiburgstrasse 15, CH-3010, Bern, Switzerland
-
Affiliation as printed
Department of Ophthalmology, Kantonsspital St. Gallen, Switzerland
-
Affiliation as printed
Department of Ophthalmology, Kantonsspital St. Gallen, Switzerland
-
Chang Gung Memorial Hospital · Taoyuan Chang Gung Memorial Hospital · Linkou Chang Gung Memorial Hospital
Affiliation as printed
Department of Ophthalmology, Linkou Main Branch, Chang Gung Memorial Hospital, Taoyuan, Taiwan
-
Chang Gung Memorial Hospital · Taoyuan Chang Gung Memorial Hospital · Linkou Chang Gung Memorial Hospital
Affiliation as printed
Department of Ophthalmology, Linkou Main Branch, Chang Gung Memorial Hospital, Taoyuan, Taiwan
-
Hanusch Hospital · Vienna Institute for Research in Ocular Surgery
Affiliation as printed
Department of Ophthalmology, Hanusch Hospital, Vienna Institute for Research in Ocular Surgery (VIROS), Vienna, Austria
-
Hanusch Hospital · Vienna Institute for Research in Ocular Surgery
Affiliation as printed
Department of Ophthalmology, Hanusch Hospital, Vienna Institute for Research in Ocular Surgery (VIROS), Vienna, Austria
-
Affiliation as printed
Ophthalmology Department, Ain Shams University, Cairo, Egypt; Watany Eye Hospital, Cairo, Egypt; Oxford Eye Hospital, Oxford University Hospitals NHS Foundation Trust, Oxford, United Kingdom
-
Affiliation as printed
Ophthalmology Department, Ain Shams University, Cairo, Egypt
-
Affiliation as printed
Department of Ophthalmology, Ankara University Faculty of Medicine, Ankara, Turkey
-
Affiliation as printed
Department of Ophthalmology, Ankara University Faculty of Medicine, Ankara, Turkey
-
University of Southern Denmark
Affiliation as printed
Department of Ophthalmology, Rigshospitalet, Glostrup, Denmark; Department of Regional Health Research, University of Southern Denmark, Odense, Denmark; Department of Ophthalmology, University Hospital of Southern Denmark, Sønderborg, Denmark
-
Fondation Asile des Aveugles · University of Lausanne
Affiliation as printed
Department of Ophthalmology, University of Lausanne, Jules-Gonin Eye Hospital, Fondation Asile des Aveugles, Lausanne, Switzerland
-
Fondation Asile des Aveugles · University of Lausanne
Affiliation as printed
Department of Ophthalmology, University of Lausanne, Jules-Gonin Eye Hospital, Fondation Asile des Aveugles, Lausanne, Switzerland
-
Fondation Asile des Aveugles · University of Lausanne
Affiliation as printed
Department of Ophthalmology, University of Lausanne, Jules-Gonin Eye Hospital, Fondation Asile des Aveugles, Lausanne, Switzerland
-
Affiliation as printed
Department of Ophthalmology, University of Lausanne, Jules-Gonin Eye Hospital, Fondation Asile des Aveugles, Lausanne, Switzerland; Department of Surgical Sciences, University of Torino, Italy
-
Leiden University Medical Center
Affiliation as printed
Department of Ophthalmology, Leiden University Medical Center, Leiden, The Netherlands
-
Camiel J. F. Boon Aachen
Leiden University Medical Center · University of Amsterdam
Affiliation as printed
Department of Ophthalmology, Leiden University Medical Center, Leiden, The Netherlands; Department of Ophthalmology, Amsterdam UMC, University of Amsterdam, Amsterdam, The Netherlands
-
Affiliation as printed
Department of Ophthalmology, Leiden University Medical Center, Leiden, The Netherlands; Department of Vitreoretinal Surgery, Rotterdam Eye Hospital, Rotterdam, The Netherlands
-
Affiliation as printed
Department of Ophthalmology, Inselspital, Bern University Hospital, University of Bern, Freiburgstrasse 15, CH-3010, Bern, Switzerland; Bern Photographic Reading Center, Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland
-
Moorfields Eye Hospital NHS Foundation Trust · University Hospital of Bern · University College London
Affiliation as printed
Department of Ophthalmology, Inselspital, Bern University Hospital, University of Bern, Freiburgstrasse 15, CH-3010, Bern, Switzerland; Moorfields Eye Hospital, NHS Foundation Trust, London, UK; Ophthalmic Surgical Intelligence Lab, London, UK; University College London, Institute of Ophthalmology, London, UK