Magnetic Resonance Imaging radiomics model predicted choroidal invasion in intraocular retinoblastoma with AUROC 0.86.
Radiomics to predict high risk features in intraocular retinoblastoma.
Aparna Ramasubramanian … Ron Korn
Eye (London, England) · 2026
Background
Detecting high risk characteristics in intraocular retinoblastoma is important for determining metastatic risk.
Methods
In this pilot study, preoperative Magnetic Resonance Imaging (Magnetic Resonance Imaging) radiomics data were compared to high-risk histopathology features after enucleation.
n = 31 eyes
Results
0.86
AUROC of 0.86 indicates strong accuracy predicting choroidal invasion; 1.0 is perfect, 0.5 is chance.
n = 31 eyes
More results
Among 31 eyes of 29 patients, 14 eyes had optic nerve invasion and 6 had choroidal invasion.
More results
The predictive model for optic nerve (optic nerve) invasion achieved Area under the receiver operating characteristic curve (AUROC) of 0.85 (accuracy 0.84, sensitivity 0.93, specificity 0.76) using the radiomic features of maximum 2D diameter slice on the unaffected optic nerve (optic nerve), correlation of grey level co-occurrence matrix (GLCM) on the affected optic nerve, and kurtosis of affected optic nerve.
“
Conclusion
Radiomics could be useful to guide the decision of globe salvage versus enucleation in advanced retinoblastoma. Larger studies are required to validate and enhance this predictive model.