Prediction of Anemia and Estimation of Hemoglobin Concentration Using a Smartphone Camera

This is a prospective convenience sample study conducted in Emergency Department (ED) patients of an academic teaching hospital.  In an algorithm derivation phase, images of both conjunctiva were obtained from 142 patients in Phase 1 using a smartphone in addition to their measured hemoglobin via venipuncture. A region of interest targeting the palpebral conjunctiva was selected from each image.  Image-based parameters were extracted and used in stepwise regression analyses to develop a prediction model of estimated hemoglobin from conjunctiva imaging (HBc). These data are in "eye_long_modeling 08JUN2021". In Phase 2, a validation model was constructed using data from 202 new ED patients, and their true hemoglobin values, found in file "clinical_set_lo_validation 08JUN2021". The final model based on all 344 patients was tested for accuracy in anemia and transfusion thresholds. Hemoglobin concentration ranged from 4.7 to 19.6 g/dL (mean 12.5).