Summary
Doctors use different computer models to predict diabetic eye disease, but these models often need different types of information and work in different ways. The authors built a web system called DR-LabStack that connects these models under one user-friendly interface so clinicians can easily use any of them. The system handles different types of data inputs and brings the models’ results together in a common format. They tested the system to make sure the models load properly and the interface works as expected. This work focuses on the software design and functionality, not yet on how well doctors find it usable or helpful in clinical practice.
What this means in practice
- •For clinical software developers: Integrate multiple pretrained diabetic retinopathy prediction models into a single web application with unified data input and output handling.
- •For medical informatics teams: Deploy a reusable web system framework to make heterogeneous clinical prediction models accessible via a common interface.
Abstract
Pretrained diabetic retinopathy (DR) prediction models differ in their input fields, serialization formats, preprocessing requirements, and output semantics. Making these models accessible through a common clinical interface therefore requires explicit coordination between the user interface and the inference service. We designed and implemented DR-LabStack, a React-Flask web system integrating four externally developed pretrained models: RuleFit, Pruned RuleFit, Elaborative XGBoost, and Two-level Ensemble. A shared form retrieves ordered model features, renders model-specific numerical and categorical controls, and constructs a positional input vector. Backend adapters load heterogeneous artifacts and apply the ensemble's accompanying scaler, while a common JSON response supports binary classification display alongside method and source information. Functional evaluation on September 8, 2026 used copied application files and real model artifacts in a documented isolated environment. All four models loaded and exposed their 14-, 6-, 8-, and 25-field contracts. Sixty-two Flask test-client requests characterized service behavior; 12 limited-vector checks confirmed invocation-path and threshold consistency. Twenty-four browser-component scenarios with mocked transport verified input ordering and result rendering and characterized input-validation behavior. The resulting system demonstrates a reusable interaction and serving workflow for heterogeneous DR models. The contribution is web-system design, integration, and software functionality; clinical effectiveness and clinician usability require separate evaluation.