ABSTRACT
Anemia is a potentially fatal disease that disturbs individuals worldwide. Conjunctival/fingertip pallor may be quantified and used to detect anemia using digital pictures acquired with a camera or a simple smartphone. Super-resolution image reconstruction can augment the spatial resolution of captured imageries, allowing for rich feature extraction based on color and texture, which is a prerequisite for hemoglobin measurement. According to WHO guidelines, an artificial neural network classifier is employed to link the hemoglobin level to be tested with the values of the quantity determined by the conventional technique. The sensitivity and specificity of anemia detection, critical for reliable anemia diagnosis, characterize the diagnostic significance of conjunctival examination in the health center and low-resource settings. The outcomes demonstrate that the proposed basic artificial neural network system can assist medical practitioners in detecting Hb in noninvasive methods in low-resource clinics. The suggested approach is suitable for identifying mild to moderate anemia. Employing super-resolution to increase deep feature extraction expands the accuracy of noninvasive Hb recognition and quantification in low-resource situations, one of the study's strengths. The pale appearance of the palpebral conjunctiva indicates anemia.
