Deoxygenation-free digital holographic microscopy discriminates sickle-cell disease in a piloted Colombian cohort via phase and shape metrics

Rapid, reagent-free screening tools for sickle-cell disease (SCD) are urgently needed in low-resource regions where carrier frequency is high and laboratory infrastructure is scarce. We demonstrate digital holographic microscopy (DHM) as a label-free SCD tool in Colombia, imaging 885 individual erythrocytes (443 cells from venous blood of six SCD patients and 442 cells from 5 healthy donors). Two direct optical observables were extracted: phase difference (∆φ) relative to the surrounding medium and projected circularity index (CI), defined as the aspect ratio between the erythrocyte’s minor and major axes. Healthy cells are characterized by ∆φ = 1.72 ± 0.30 (mean ± std) rad and CI = 0.94 ± 0.04, whereas SCD cells are distributed around ∆φ = 3.00 ± 0.96 rad and CI = 0.70 ± 0.19. Despite some minimal overlap between both distributions, SCD cells are statistically distinguishable from healthy ones based on both ∆φ and CI (p < 0.001). A simple rule (∆φ > 2.2 radians ∧ CI < 0.85) classified individual cells with 97.40% accuracy, 94.81% sensitivity, and 100% specificity. While sophisticated classifiers can further optimize boundaries and enable anemia sub-typing, the distinctive contribution of this work is its deliberately minimalist rule – requiring no model training or engineered features – offering immediate and interpretable triage well-suited for low-resource workflows.