OphJev: endpoint-aware System One decisions for ophthalmic imaging and clinical text
- 1Tsinghua University
Abstract
Ophthalmic decisions require probabilities tied to a clinical question, a candidate set and an operating policy. We present OphJev, an endpoint-aware System One framework that combines frozen image and text representations, candidate scoring, schema-specific calibration and selective generalist queries for fundus and OCT imaging. Across eight external resources, multiple representation families, three generalist sizes and three seeds, controlled interventions separate the effects of wording, supervision and computation. Schema-aware affine calibration lowers mean held-out Brier loss and selective risk in all eight primary image-text configurations. On HYGD, matched-supervision fusion raises AUROC from 0.772 to 0.798, while a separate gain-routing policy lowers Brier from 0.618 to 0.599 using at most 20% logical generalist image calls. These effects demonstrate complementary decision-level capabilities rather than universal gains: external transfer and larger specialist banks remain cohort dependent. A retrospective text extension evaluates 1,996 clinical notes and 91 ophthalmic questions; validation-based calibration improves note probability scores, while a lexical control remains strong. OphJev provides an auditable decision framework; prospective clinical benefit remains unvalidated.
Related
- Artificial intelligence for ophthalmic vessel segmentation: from multimodal imaging to clinical translation
- Development and validation of an AI-enhanced prediction model for 3-year visual decline in patients with diabetes using ophthalmic imaging: protocol for a real-world longitudinal cohort study
- Therapies for Alcohol Use Disorder in Alcohol-associated Hepatitis Clinical Trials: Integrated Trial Design with a Composite Clinical Outcome as the Primary Endpoint
- inClinico Agent for clinical trial endpoint prediction and prospective evaluation
- The integration of artificial intelligence into computed tomography: from segmentation to the standardisation of clinical decisions
BibTeX
@misc{du2026ophjev,
title = {{OphJev: endpoint-aware System One decisions for ophthalmic imaging and clinical text}},
author = {Z. Du and Z. Deng and L. Ma},
howpublished = {medRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.03.26364645},
url = {https://www.medrxiv.org/content/10.64898/2026.10.03.26364645v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - OphJev: endpoint-aware System One decisions for ophthalmic imaging and clinical text
AU - Z. Du
AU - Z. Deng
AU - L. Ma
PY - 2026
DA - 2026/10/03
PB - medRxiv
DO - 10.64898/2026.10.03.26364645
UR - https://www.medrxiv.org/content/10.64898/2026.10.03.26364645v1
AB - Ophthalmic decisions require probabilities tied to a clinical question, a candidate set and an operating policy. We present OphJev, an endpoint-aware System One framework that combines frozen image and text representations, candidate scoring, schema-specific calibration and selective generalist queries for fundus and OCT imaging. Across eight external resources, multiple representation families, three generalist sizes and three seeds, controlled interventions separate the effects of wording, supervision and computation. Schema-aware affine calibration lowers mean held-out Brier loss and selective risk in all eight primary image-text configurations. On HYGD, matched-supervision fusion raises AUROC from 0.772 to 0.798, while a separate gain-routing policy lowers Brier from 0.618 to 0.599 using at most 20% logical generalist image calls. These effects demonstrate complementary decision-level capabilities rather than universal gains: external transfer and larger specialist banks remain cohort dependent. A retrospective text extension evaluates 1,996 clinical notes and 91 ophthalmic questions; validation-based calibration improves note probability scores, while a lexical control remains strong. OphJev provides an auditable decision framework; prospective clinical benefit remains unvalidated.
ER -