Imaging the respiratory effects of truncal adiposity in acute hypoxemic respiratory failure

The National Heart, Lung, and Blood Institute awarded Massachusetts General Hospital $1.356 million on September 10, 2025, under the Cardiovascular Diseases Research program (CFDA 93.837) to image the respiratory effects of truncal adiposity in acute hypoxemic respiratory failure.

The project uses computed tomography imaging and artificial intelligence-based segmentation to test the hypothesis that the magnitude and distribution of truncal adiposity governs lung collapse and severity of acute hypoxemic respiratory failure in obese patients. The research group applies novel CT image analysis involving registration and artificial intelligence-based segmentation to quantify lung collapse and injury progression. The project will evaluate how measurements of intrathoracic pressure, global respiratory mechanics using forced oscillation technique, and regional ventilation using electrical impedance tomography can detect the effects of truncal loading on lung collapse at bedside. The work leverages a ventilated swine model of acute hypoxemic respiratory failure and clinical infrastructure for monitoring patients with acute hypoxemic respiratory failure.

Performance runs through June 30, 2030, with work performed in Boston, Massachusetts.

The University of Iowa receives a sub-award (R01HL177025-GR1003372-S01) awarded January 30, 2026, to analyze and quantify computed tomography images of swine and humans in respiratory failure. A postdoctoral fellow will develop, validate, and run image analysis tools for segmentation and registration of lungs, airways, and vasculature, using manually annotated training data to train deep neural networks. Deliverables include segmentations of pulmonary structures and image registration deformation fields with derived regional biomechanical maps.

Abstract

Obesity affects more than 40% of the united states population and is a major risk factor for acute hypoxemic respiratory failure (ahrf). Ahrf patients with obesity have more ventilator-related morbidity and may be less responsive to standard respiratory care treatments, such as prone position and positive end expiratory pressure (peep) titration, compared to patients of normal body weight. Excess truncal adiposity in obesity elevates intrathoracic pressure (ITP) and compresses lung tissue. Our prior research has shown that obesity also modifies ventilator-induced lung stretch. In addition to the mass of adiposity, published and preliminary data suggest that the distribution of adiposity influences ITP and may affect the magnitude and pattern of lung collapse. Imaging the topographic effects of truncal adiposity on lung tissue may therefore 1) provide insight into ahrf severity and treatment response in obesity and 2) identify viable tools that can quantify the effects of truncal loading at bedside. Our research group has extensive expertise in applying novel computed tomography (CT) image analysis involving registration and artificial intelligence-based segmentation to quantify lung collapse and injury progression in ahrf. In this proposal, we will use CT to test our hypothesis that the magnitude and distribution of truncal adiposity governs lung collapse and severity of ahrf in obesity. Using CT as reference, we will further evaluate how measurements of ITP (esophageal manometry), global respiratory mechanics (forced oscillation technique), and regional ventilation (electrical impedance tomography) can detect the effects of truncal loading on lung collapse at bedside. To achieve the aims of this proposal, we have developed a ventilated swine model of ahrf and a robust clinical infrastructure to monitor patients with ahrf in our hospital. We will first measure in our animal model the effect of load distribution on imaging, biological, and physiological markers of ahrf progression during prolonged mechanical ventilation for 24 hours. We will then evaluate the effects of prone positioning and peep titration on the regional inflation characteristics and radiological abnormalities using sequential imaging in conditions of truncal loading. Lastly, in ventilated patients with ahrf and in healthy volunteers with and without obesity, we will measure associations between thoracic adiposity, ITP, and lung collapse in supine and prone position. Data obtained from our proposal will advance our understanding of obesity in ahrf, informing individualized clinical management to assist in the early identification and treatment based on the morphological and biomechanical characteristics of each patient.