# Evaluation of Quantitative Computed Tomography Indices in Patients with Pneumonia and Acute Respiratory Failure in the Intensive Care Unit (ICU)

## TL;DR

This study shows that CT scan measurements of lung density and volume differences can predict mortality in ICU patients with pneumonia and respiratory failure.

## Contribution

The study identifies inter-lung differences in high- and medium-density volumes as novel independent predictors of mortality in pneumonia patients.

## Key findings

- Difference for Lung HDV > 193 mL is an independent risk factor for mortality (aOR: 4.29, p = 0.041).
- In unilateral pneumonia, Difference for Lung MDV > 219 mL and HDV > 193 mL are strong mortality predictors (aOR: 9.30 and 10.85 respectively).
- Lung volume differences show strong diagnostic performance (AUC: 0.808) in predicting mortality.

## Abstract

Background: In this study, we aimed to explore the relationship between quantitative indices derived from computed tomography (CT) attenuation histograms and disease prognosis in patients with pneumonia and acute respiratory failure. We also sought to assess the effectiveness of these parameters as clinical prognostic markers. Methods: CT images of patients with pneumonia and acute respiratory failure were analyzed using Vitrea® Advanced Visualization software. The analyzed quantitative CT (qCT) indices included mean lung Hounsfield unit (HU) and density-based volume measurements, specifically low-, medium-, and high-density volume (LDV, MDV, and HDV). Comparative analyses were performed to examine the differences in the volume density between the lungs bilaterally; these were accompanied by regional analyses and density indices. All indices were calculated using previously defined and validated Hounsfield unit (HU) thresholds, which helped to ensure accurate and consistent quantitative measurements and facilitated a more robust evaluation of the prognostic potential of qCT parameters. Results: Quantitative CT indices proved to have significant prognostic value in predicting mortality. In multivariable analysis, Difference for Lung HDV > 193 mL emerged as an independent risk factor (aOR: 4.29, p = 0.041). The prognostic significance was especially evident in patients with unilateral dominant pneumonia, where Difference for Lung MDV >219 mL (aOR: 9.30, p = 0.03) and Difference for Lung HDV > 193 mL (aOR: 10.85, p = 0.02) emerged as strong independent predictors of mortality. In this subgroup, lung volume differences demonstrated the strongest diagnostic performance (AUC: 0.808, 95% CI: 0.667–0.908, p < 0.001). Conclusions: Clinical outcomes are associated with quantitative CT-derived lung volume and density difference indices. Inter-lung differences in Lung MDV and Lung HDV are linked to mortality and may provide additional prognostic information beyond conventional imaging methods. Prospective studies should be conducted to validate these findings, and caution should be exercised during their interpretation.