Non-apnea Saline-contrast EIT for Lung Perfusion Imaging in Spontaneously Breathing Patients

Study DesignThis is a prospective, randomized, cross‐over, within‐subject validation study ofaccuracy.The primary objective is to quantitatively assess the agreement between lung perfusiondata acquired during uninterrupted spontaneous breathing (non‐apnea method) and dataacquired during the standard end‐expiratory pause (pause method).Participant Population and Respiratory SupportThe study enrolls stable adult patients with active spontaneous breathing efforts.Participants are categorized by the level of respiratory support they receive:Conventional Oxygen Therapy (Nasal Cannula), High‐Flow Nasal Oxygen (HFNO), or PressureSupport Ventilation (PSV).Conventional Oxygen Therapy includes patients using a nasal cannula or simple maskwithout pressure support.HFNO and PSV groups include patients receiving constant high‐flow oxygen orpatient‐triggered mechanical ventilation.Procedures and InterventionsEach participant receives two 10 mL boluses of 10% NaCl (hypertonic saline) via a centralvenous catheter.A 10‐minute washout period is maintained between injections.In the Pause Method, injection occurs during a 5‐8 second end‐expiratory hold (coachedfor nasal cannula/HFNO; ventilator‐controlled for PSV).In the Non‐Apnea Method, injection occurs during continuous, ongoing breathing withoutany breathing commands or ventilator setting changes.Technical Approach to Ensure AccuracySpecialized algorithms are used to extract saline‐contrast signals from the "noise" ofspontaneous tidal breaths.Advanced frequency filtering is applied to separate ventilation‐related andperfusion‐related impedance changes.In non‐apnea mode, the software identifies the perfusion window using a pixel‐triggeredapproach.This method detects the impedance drop relative to the End‐Expiratory Lung Impedance(EELI) baseline to maintain precision despite moving baselines.Outcome MeasuresThe primary outcome is the pixel‐wise correlation of perfusion maps between the twomethods, assessed by the Spearman correlation coefficient (r).Secondary outcomes include the agreement of V/Q match%, dead space%, and shunt% usingBland‐Altman analysis.The study also assesses whether the type of support (Nasal Cannula vs. HFNO vs. PSV)affects the accuracy of the non‐apnea method.Feasibility is documented by the success rate of analyzable acquisitions and categorizedfailure reasons.