Related Experiment Video
Updated: Sep 28, 2026

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
Unilateral Orbital Lymphomas: A Structured Multidisciplinary Workflow for Diagnosis and Management
Maria Angela Romeo1, Alessandro Gaeta2, Giuseppe Giannaccare3
1Department of Ophthalmology, University Magna Graecia of Catanzaro, Catanzaro, Italy.
Purpose:
To describe a structured multidisciplinary workflow for the diagnosis and management of unilateral orbital lymphoma and to illustrate its clinical application through representative cases from a tertiary referral practice. Despite advances in orbital imaging and hematopathology, no standardized framework currently integrates clinical assessment, quantitative MRI parameters, immunohistochemical subtyping, and systemic staging into a reproducible diagnostic-therapeutic pathway.
Patients And Methods:
A stepwise workflow integrating clinical assessment, MRI with diffusion-weighted imaging (DWI) and apparent diffusion coefficient (ADC) measurement, histopathologic confirmation with immunohistochemistry (IHC), and 18F-FDG PET/CT staging was developed at a tertiary referral center through multidisciplinary consensus. The workflow was applied to six patients with biopsy-confirmed unilateral orbital lymphoma (marginal zone, n=3; follicular, n=2; lymphoplasmacytic, n=1), selected to represent the clinical and radiologic spectrum encountered in practice.
Results:
The workflow facilitated systematic evaluation across all cases. MRI consistently demonstrated characteristic features: well-circumscribed, homogeneously enhancing lesions molding to orbital structures, with mean ADC values of 0.55-0.70 × 10-3 mm2/s, supporting early biopsy decisions. IHC confirmed subtype-specific immunophenotypes guiding treatment selection. PET/CT identified systemic involvement in two patients (33%), modifying the therapeutic approach. Localized disease achieved remission with low-dose radiotherapy; disseminated cases responded to rituximab-based regimens. At a mean follow-up of 27 months, all patients were alive, with four achieving a complete response.
Conclusions:
The proposed workflow provides a reproducible framework for integrating clinical, radiologic, and histopathologic data in orbital lymphoma evaluation. Systematic application may support diagnostic accuracy, facilitate individualized staging, and inform evidence-based treatment selection, although prospective multicenter validation is needed before broader adoption. This methodology complements a recently accepted companion paper on unilateral orbital lesions, together providing a comprehensive resource for orbital disease assessment.
