Time-Domain Analysis of Chest Impedance for Venous Air Embolism Detection: An Open-Source Software Tool and Single-Cadaver Pilot Study.

Venous air embolism remains one of the leading intraoperative neurosurgical complications and causes of intraoperative neurosurgical death. While transesophageal echocardiogram and precordial Doppler are commonly used in neurosurgical cases to detect entrained air, both have significant limitations, and neither tool has been adapted to automated notification. We theorize that (1) higher frequency variations (relative to the respiratory cycle), and in severe circumstances, (2) relatively non-varying increases in chest impedance can detect venous air emboli using non-invasive or minimally invasive continuous techniques. This would allow for automated reporting of these findings, which would be of use to anesthesia providers. The current tools (primarily echocardiogram and precordial Doppler) lack the ability to provide automated or centralized alarm functions and can be technically difficult, more costly to perform, and provide non-continuous information. A proof-of-concept open-source software tool has been developed to allow for collaborative development of this technique, and preliminary testing has been performed on a previously described cadaveric model for venous air embolism, with a single cadaver as a pilot study.