Noise effect analysis and pulmonary perfusion estimation in electrical impedance tomography

Area of Science:

• Medical Imaging
• Physiological Monitoring
• Critical Care Medicine

Background:

• Electrical Impedance Tomography (EIT) is a non-invasive bedside monitoring tool for mechanically ventilated patients.
• Accurate estimation of pulmonary perfusion is crucial for managing respiratory conditions.

Purpose of the Study:

• To evaluate pulmonary perfusion estimation using EIT during a hypertonic saline bolus procedure.
• To assess the impact of signal drift and cardiac partial volume effects on EIT outcomes.

Main Methods:

• Utilized first-pass kinetics modeling on EIT data to differentiate lung and hybrid pixels.
• Analyzed the influence of signal drift and cardiac partial volume effects on perfusion metrics.
• Validated model performance with both simulated and real patient data.

Main Results:

• Uncompensated signal drift led to a 33% overestimation of the maximum slope.
• Uncompensated cardiac partial volume effect caused a 13.9% underestimation of the maximum slope.
• The developed model demonstrated performance within physiologically feasible limits.

Conclusions:

• EIT-based pulmonary perfusion estimation is feasible but sensitive to noise.
• Signal drift and cardiac effects require compensation for accurate EIT analysis in ventilated patients.
• This study highlights the importance of noise mitigation in EIT for reliable bedside monitoring.