Researchers have developed SpecF2M, a novel multi-task network designed to estimate axial length and refractive error from pediatric fundus photographs. This spectral-aware network integrates an anatomy-guided enhancement module and a hybrid spatial-spectral backbone to estimate components like axial length, sphere, and cylinder. In a study involving over 4,000 child visits and nearly 7,000 fundus images, SpecF2M demonstrated superior performance compared to standard CNN and ViT baselines for axial length and sphere estimation, achieving mean absolute errors of 0.5347 mm and 0.7062 D, respectively. The findings suggest that fundus photography can be used for screening pediatric myopia indicators, though external validation is needed before clinical deployment. AI
IMPACT This model could enable more accessible and cost-effective screening for pediatric myopia and related eye conditions.
RANK_REASON The cluster describes a new AI model presented in an academic paper for a specific medical imaging task. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
