Inference of UV-VIS Characterization Result of Human Blood Hemoglobin Levels Using Artificial Neural Network (ANN)

Various studies aim to eliminate the reagent need for blood haemoglobin level measurement. This study aims to test the Artificial method Neural Network (ANN) to estimate blood haemoglobin levels without using reagents. The results of UV-Vis spectrophotometry measurements in the previous study were used as secondary data with a wavelength range of 300nm-500 nm and processed using an Artificial Neural Network (ANN) inference engine to test the effectiveness. So that can be categorize the best wavelength range candidates based on the highest absorbance and standard deviation, which are in the range of 340 and 360, 420 nm and 440 nm, 480 nm and 500 nm. Accuracy tests also result in the same wavelength candidates based on absorbance and standard deviation. It can be concluded that the ideal wavelength range for monitoring blood haemoglobin levels in humans, both in terms of accuracy and diagnostically, can be carried out in the wavelength range of around 420 nm to 500 nm.