Non-invasive quantification of total hemoglobin concentrations by visible-light spectroscopic optical coherence tomography: a validation study on healthy volunteers.

Abstract

Clinically, a high demand exists for the noninvasive assessment of total hemoglobin concentrations in vulnerable patient groups. Despite the sensitivity of many optical modalities to the absorption of hemoglobin, it has remained challenging to optically measure hemoglobin concentrations noninvasively with sufficient accuracy. Compared to other optical modalities, visible-light spectroscopic optical coherence tomography (vis-sOCT) has the unique ability to quantify optical properties within highly confined tissue volumes. However, total hemoglobin quantification in vivo is still limited due to a lack of dedicated acquisition schemes, processing methods, and validation against a gold standard. In this work, we introduce an approach that combines optical angiography with spatially resolved visible-light spectroscopy to obtain clinically representative estimations of the total hemoglobin concentration in the skin microcirculation. We validate our approach in vivo on a group of 27 healthy volunteers against hemoglobin measurements on venous blood samples with a commercial blood analyzer. The measured hemoglobin concentration by vis-sOCT correlates moderately, but significantly with the blood analyzer (R = 0.47, p = 0.015). This approach quantifies hemoglobin with a mean bias of 0.6 g dL^-1 and an average standard deviation of 1.3 g dL^-1 in the healthy hemoglobin range (14-19 g dL^-1). Moreover, the results highlight the known influence of gender on the comparison between the microcirculatory and venous hemoglobin concentration. This observed influence of gender further suggests that vis-sOCT is sensitive to microcirculatory changes in hemoglobin concentration.