Truace
Subscribe

Interpretable Machine Learning for Course Prediction in Central Serous Chorioretinopathy Using Baseline OCT from a Multicenter Cohort

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 retrospective…

By the Truvace desk1 min read1 sourcePeer-reviewedTRV-2026-1627· cite & verify
Interpretable Machine Learning for Course Prediction in Central Serous Chorioretinopathy Using Baseline OCT from a Multicenter Cohort
"CRE-003 OphaGo - Retinal Imaging Adapter - Huced Despro ITS" by Djoko Kuswanto is marked with CC0 1.0. To view the terms, visit https://creativecommons.org/publicdomain/zero/1.0/.

In brief

Researchers built an interpretable Random Forest model using retinal measurements pulled automatically from the first OCT scan of 329 untreated eyes with central serous chorioretinopathy. Eyes were later labeled acute or chronic depending on whether subretinal fluid resolved during follow-up averaging 4.2 years.

The findings suggest that early scans already carry signals about who is likely to develop persistent disease, which could shape counseling and risk identification. The work is retrospective, so it shows predictive association in past records rather than proof that using the tool improves patient outcomes.

Main points

  1. The cohort covered 329 treatment-naïve eyes from 291 patients across eight international centers.
  2. Of the eyes, 92 (28.0%) were acute and 237 (72.0%) were chronic CSC.
  3. Chronic CSC patients were significantly older, at 51.7 ± 12.6 versus 43.2 ± 9.6 years.
  4. Features included outer nuclear layer and PR+RPE thickness, SRF and PED volumes, and biomarker probabilities across ETDRS regions.
  5. SHAP values were used to rank biomarker importance in the classifier.

The gain

Automatically detected OCT biomarkers at first presentation predicted whether treatment-naïve CSC eyes would become chronic, which may support prognosis and patient counseling.

The rundown

The study used an AI-based platform, the Discovery OCT Biomarker Detector from RetinAI AG, to quantify retinal layer thicknesses and fluid volumes in central, parafoveal, and perifoveal regions from baseline scans.

Chronicity was defined by whether subretinal fluid resolved on follow-up. Model discrimination was assessed with area under the receiver operating characteristic curve, and SHAP values explained which biomarkers drove predictions.

The authors conclude that markers of outer retinal integrity, fluid configuration, and RPE dysfunction may improve prognostication and help identify individuals at increased risk of persistent CSC.

Sources

  1. Peer-reviewedAmerican Journal of Ophthalmology2026-10-08

The debate