Simple method for estimation of hemoglobin in human blood using color analysis

Area of Science:

• Biomedical Engineering
• Clinical Chemistry
• Artificial Intelligence in Medicine

Background:

• Anemia is a significant global health issue, particularly affecting women and children.
• The cyanmethemoglobin method is the WHO-recommended standard for hemoglobin (Hb) measurement.
• Colorimetric methods offer approximate Hb results and are recommended by the WHO for low-resource settings, but are prone to human interpretation errors.

Purpose of the Study:

• To evaluate an artificial neural network (ANN) approach for estimating hemoglobin (Hb) levels in human blood.
• To address the limitations of subjective interpretation in color-measurement techniques for Hb estimation.
• To develop a potentially more accurate and objective method for Hb measurement in resource-limited environments.

Main Methods:

• An artificial neural network (ANN) model was developed and trained using 2007 blood samples.
• The ANN utilized color-coded values of blood samples as input.
• Hemoglobin (Hb) values, determined by the cyanmethemoglobin method, served as the desired output for training.

Main Results:

• The study demonstrated a strong correlation between the color of blood samples and their corresponding hemoglobin (Hb) levels.
• The ANN approach showed promise in accurately estimating Hb concentrations based on visual color data.
• This suggests that objective color analysis can be a viable proxy for quantitative Hb measurement.

Conclusions:

• Artificial neural networks (ANNs) can effectively estimate hemoglobin (Hb) levels by analyzing blood sample color.
• ANNs offer a potential solution to overcome human interpretation errors in color-based Hb measurement methods.
• This technology could enhance anemia diagnosis and management, especially in low-resource settings worldwide.