ANALYSIS OF HEMOGLOBIN LEVELS USING IMAGE-BASED METRICS: A MULTIVARIATE APPROACH

Abstract

This study investigates the relationship between hemoglobin (Hb) levels and image analysis metrics derived from blood sample images. Utilizing a dataset of 425 observations, we employed descriptive statistics, graphical analysis, confidence intervals, hypothesis testing, and multivariate regression to explore the impact of various image-based features on hemoglobin levels. Our findings reveal that blue channels in the image analysis have the most significant influence on hemoglobin levels, with a strong negative correlation. The study provides insights into the potential use of image-based metrics for hemoglobin analysis and highlights the importance of considering color channels in medical diagnostics.