Simplifying Bedside Lung Perfusion Imaging Using Electrical Impedance Tomography - PROJECT SUMMARY Acute Respiratory Distress Syndrome (ARDS) is a life-threatening condition characterized by widespread pulmonary inflammation that leads to lung injury. In severe cases, the mortality rate can exceed 40%. Imaging provides a topographical distribution of perfusion, which is typically heterogeneous in ARDS. This information is crucial for developing treatment strategies. However, assessing lung perfusion in ARDS is challenging due to a lack of suitable bedside measuring tools. Electrical Impedance Tomography (EIT) is a relatively new modality that shows promise as a bedside tool for measuring subtle changes in lung perfusion, thereby eliminating the need to transport these critically ill patients to offsite imaging facilities. The current EIT acquisition, involving hypertonic contrast administered via a central venous line during a 30-second breath-hold, is feasible only for sedated patients who can tolerate interruptions to their respiratory cycle and have a central line in place. However, central venous catheterization has associated risks, prompting a shift toward the use of peripheral intravenous catheters. Furthermore, recent ARDS definitions include non-mechanically ventilated patients, underscoring an urgent need for EIT methods that do not require the cessation of spontaneous breathing. Currently, acquisition methods that bypass these limitations remain undeveloped. Our study aims to advance EIT application by adapting it for use during regular breathing cycles in both fixed- rate and spontaneous (Aim 1), and utilizing peripheral venous injections (Aim 2), thereby eliminating the need for induced apnea and central line placement. We also seek to optimize image acquisition and strengthen post- analysis techniques by leveraging successful strategies from more established dynamic contrast-enhanced (DCE) MRI and CT methods and applying them to EIT. This will be achieved through animal experiments using equal numbers of male and female Yorkshire pigs under healthy and lung-injured conditions, as well as observational studies in 20 ARDS patients with pneumonia. Key methodological innovations include the development of a model-based filter to separate the enhancement trace from contrast injections and decouple it from cyclic respiratory patterns, facilitating the capture of perfusion data without interrupting normal respiration. We plan to explore deconvolution methods and compare traditional first-pass kinetic models to assess their resilience to lower SNR caused by peripheral injections and hypoperfusion due to injury. Additionally, we will investigate analysis beyond the first-pass kinetics to address interstitial pathology. The success of this study could broaden the application of perfusion measurement, streamline the process, and facilitate safer, more frequent, and precise bedside evaluations of lung perfusion in ARDS patients, ultimately enhancing patient care and clinical outcomes.
