Background: Targeting systemic hemodynamic parameters in critical care settings does not always improve patient outcomes. The cellular oxygen metabolism (COMET) monitor noninvasively measures mean mitochondrial oxygen tension (mitoPO₂) and its variance via delayed fluorescence of protoporphyrin IX. However, microvascular oxygenation is often heterogeneous, allowing hypoxic and normoxic tissue to coexist, potentially leading to organ dysfunction. 

Methods: A new algorithm was developed to study underlying mitoPO₂ distributions. To evaluate performance, the algorithm was first tested on simulated unimodal and bimodal mitoPO₂ distributions across varying signal-to-noise ratios (SNRs). It was then applied to clinical data from cardiac surgery patients.

Simulation results: The algorithm accurately recovered the mean mitoPO₂  and underlying distributions across varying SNRs.

Clinical results: A total of 47 patients were included in the analysis, with 28 developing cardiac surgery-associated acute kidney injury (CSA-AKI) and 19 not. The mitoPO₂ calculated by the developed algorithm generally followed the results from the COMET closely but showed overestimation in the lower oxygen range. The CSA-AKI group spent significantly more intraoperative time with mitoPO₂ <25mmHg consistent with previous research.

Conclusion: The developed algorithm proved to be reliable in estimating the mean mitoPO₂. Larger patient cohorts are needed to validate the role of the algorithm in CSA-AKI risk assessment.

Note: title and abstract differ from original due to confidentiality.