Microcirculatory assessment of patients under VA-ECMO.

BACKGROUND: Veno-arterial extracorporeal membrane oxygenation (VA-ECMO) is an effective technique for providing emergency mechanical circulatory support for patients with cardiogenic shock. VA-ECMO enables a rapid restoration of global systemic organ perfusion, but it has not been found to always show a parallel improvement in the microcirculation. We hypothesized in this study that the response of the microcirculation to the initiation of VA-ECMO might identify patients with increased chances of intensive care unit (ICU) survival. METHODS: Twenty-four patients were included in this study. Sublingual microcirculation measurements were performed using the CytoCam-IDF (incident dark field) imaging device. Microcirculatory measurements were performed at baseline, after VA-ECMO insertion (T1), 48-72 h after initiation of VA-ECMO (T2), 5-6 days after (T3), 9-10 days after (T4), and within 24 h of VA-ECMO removal. RESULTS: Of the 24 patients included in the study population, 15 survived and 9 died while on VA-ECMO. There was no significant difference between the systemic global hemodynamic variables at initiation of VA-ECMO between the survivors and non-survivors. There was, however, a significant difference in the microcirculatory parameters of both small and large vessels at all time points between the survivors and non-survivors. Perfused vessel density (PVD) at baseline (survivor versus non-survivor, 19.21 versus 13.78 mm/mm2, p = 0.001) was able to predict ICU survival on initiation of VA-ECMO; the area under the receiver operating characteristic curve (ROC) was 0.908 (95 % confidence interval 0.772-1.0). CONCLUSION: PVD of the sublingual microcirculation at initiation of VA-ECMO can be used to predict ICU mortality in patients with cardiogenic shock.

Tissue phantoms to compare spatial and temporal offset modes of deep Raman spectroscopy.

Time-resolved and spatially offset Raman spectroscopies have previously been demonstrated for depth analysis through strongly scattering, non-transparent materials. In this study, several series of tissue phantoms were created with varied compositions and thicknesses to compare the potential of these different Raman techniques for biomedical applications. Polydimethylsiloxane (PDMS) phantoms were made with TiO2 particles suspended as a scattering agent, mimicking the scattering properties of biological tissues. The phantom layers contained embedded biomineral simulating inclusions (sphere or layer-shaped) with varied carbonate to phosphate ratios. The tissue phantoms were studied using Time Resolved Raman Spectroscopy (TRRS), Spatially Offset Raman Spectroscopy (SORS), and their combination, using a single instrumental setup with picosecond pulsed excitation at 720 nm and two different detectors. A comparison is made of the efficiency of these techniques to resolve chemical information from these heterogeneous scattering phantom samples. Measurements with continuous wave detection were found to offer a better signal-to-noise ratio than with TRRS, and in SORS measurements ratios of target to matrix signal were found to vary depending on the structural geometry and optical properties of the phantoms. Anomalous SORS behaviour, in which the relative contribution from the target decreases with offset, was observed in cases where the target was highly scattering and the top layer was relatively transparent. Time gating with an intensified charge-coupled device (ICCD) detector can yield more direct information on the depth of the hidden material.