Abstract Body (Do not enter title and authors here): Abstract Body (2500):
Introduction: Accurate quantification of cerebral oxygenation is critical for managing children with congenital heart defects (CHD) during pediatric critical care and procedural operations. While Continuous-wave near-infrared spectroscopy (CW-NIRS) is a clinically used tool for cerebral oximetry monitoring, its quantitative accuracy, especially during hypoxemia, is limited. Advanced frequency domain diffuse optical spectroscopy (FD-DOS) offers improved quantitative accuracy by directly measuring tissue scattering properties. While differences between FD-DOS and CW-NIRS have been observed, a direct clinical comparison against gold standard measurements in the clinical setting has been lacking.

Methods: This prospective observational study enrolled children diagnosed with cyanotic CHD undergoing cardiac catheterization. Gold standard cerebral tissue oxygenation (StO2) was calculated from invasive superior vena cava (SVC) and descending aorta (DA) blood co-oximetry measurements with an invasive catheter, assuming a widely accepted tissue venous blood volume fraction of 0.75, i.e., . Cerebral tissue oxygenation was also measured during cardiac catheterization non-invasively with a CW-NIRS INVOS oximeter and a custom FD-DOS oximeter. The non-invasive measurements were performed in parallel with sensors placed on the left and right foreheads.

Results: Data from 16 patients included gold standard measurements, among which useful FD-DOS data were available in 12 subjects, and CW-NIRS data in 8 patients. The FD-DOS measurements significantly correlated with the invasive gold standard (R = 0.807, p = 0.002), but CW-NIRS measurements were not (R = -0.099, p = 0.816) (Fig. 1). Although FD-DOS underestimated the invasive standard, possibly because of a fixed venous fraction derived from healthy patients. Children recruited in this study had Glenn physiology, where the higher central venous pressure (~12 mmHg) may lead to higher venous fraction. Despite the limited CW-NIRS samples, its monitoring accuracy is known to be limited during hypoxemia (DA range 71-97% and SVC range 52-76%). For 5 shared patients, FD-DOS (R = 0.88, p = 0.049) continued to outperform CW-NIRS (R = 0.118, p = 0.850) in correlation with invasive measurements.

Conclusions: FDDOS measurements of cerebral tissue oxygenation demonstrated improved accuracy compared to commercially used CW-NIRS based technology.