Hemoglobin estimation using ultra-low path length in microfluidic chips by quantifying Soret band

Abstract

Blood is the most analyzed body fluid for diagnostic purposes, and complete blood count is a widely performed blood test, wherein hemoglobin estimation is performed colorimetrically, while other parameters including counts of platelets, Red Blood Cells (RBCs) and White Blood Cells (WBCs) are estimated using imaging or impedance or light scattering techniques. Artificial Intelligence (AI) powered automated imaging systems in conjunction with microfluidic chips are some of the most promising cost-effective medical diagnostic solutions poised to revolutionize the field of Point-of-Care (POC) healthcare. Performing imaging based colorimetry would enable minimizing the cost and also the footprint of POC blood analyzers. We report the development and verification of an imaging based on-chip colorimetric assay to estimate hemoglobin in blood using ultra-low path lengths by transitioning from a widely utilized Q band absorbance peak to a more intense Soret peak associated with Sodium Lauryl Sulfate (SLS) assay. Initial SLS reagent customization characterization of the SLS reagent was performed using a nanospectrophotometer with in-house prepared hemoglobin standards and also whole blood samples. Subsequently, the imaging based SLS assay was optimized on a microfluidic chip (chamber height: 190 µm) in conjunction with an automated microscope (AI-100) equipped with a violet LED whose emission peak coincided with the Soret peak of the SLS–hemoglobin complex. The thus optimized on-chip hemoglobin assay was verified for clinical performance using a sample set consisting of 30 blood samples. The verification study indicated an accuracy (R2) of 0.98, a sensitivity of 100%, and a specificity of 87.5%. Overall, we present an imaging based on-chip hemoglobin assay using a microfluidic chip with ultra-low path lengths by quantifying the Soret peak associated with the customized SLS assay.