Estimating the refractive index of oxygenated and deoxygenated hemoglobin using genetic algorithm - support vector...

Area of Science:

• Biophysics
• Computational Biology
• Hematology

Background:

• The refractive index of hemoglobin is crucial in hematology, correlating with disease pathophysiology.
• Measuring hemoglobin's refractive index is challenging due to absorption and measurement artifacts, leading to inconsistent values.
• Existing methods are time-consuming, laborious, and expensive.

Purpose of the Study:

• To develop a rapid and accurate computational intelligent approach for estimating the real part of the refractive index of hemoglobin.
• To utilize Genetic Algorithm/Support Vector Regression (GA-SVR) models for this estimation.

Main Methods:

• Experimental data on wavelengths and hemoglobin concentrations were used.
• Highly accurate GA-SVR models were built using this experimental data.

Main Results:

• The developed GA-SVR methodology demonstrated high accuracy with low root mean square errors (4.65 × 10⁻⁴ for oxygenated, 4.62 × 10⁻⁴ for deoxygenated hemoglobin).
• Models showed excellent correlation coefficients (r = 99.85% for oxygenated, r = 99.84% for deoxygenated hemoglobin), confirming strong agreement between predicted and experimental results.

Conclusions:

• The proposed GA-SVR models are accurate and relatively simple to implement.
• These models are expected to serve as valuable references for future research on the optical properties of blood.