Quantitative
Chest CT Analysis in Relationships between CT Patterns, Virus Load, and
Pathophysiological States in SARS-CoV-2 infected Patients: ACross-Sectional
Observational Study
Wang Yan1*, Meng Haining2*, Wang
Sumei1, Jia Chao1, Lian
Zhiyuan1, Xie Weifeng1, Qu
Yan1#
1.Department of Intensive Care Unit, Qingdao Hospital, University of
Health, and Rehabilitation Sciences (Qingdao Municipal Hospital).
Address: No.5 Zhonghai East Road, Southern District, Qingdao City,
Shandong Province, China.
Wang Yan:
E-mailkeliven@163.net
Qu Yan E-mail:qdquyan@aliyun.com #Qu Yan is
the corresponding author.
2.Department of Emergency Medicine, Medical College of Qingdao
University, Qingdao, Shandong, China;
Meng Haining: E-mailmenghaining0307@163.com
*These authors contributed to the work equllly and should be regarded as
co-first authors.
Abstract
CT imaging is often used to confirm COVID-19, playing a crucial role in
the diagnosis and assessment due to its high sensitivity. The purpose of
this study is to investigate results of quantitative CT analysis for CT
patterns in SARS-CoV-2 infected patients, and how these relate to viral
load and pathophysiological states.
We
recruited patients who had confirmed SARS-CoV-2 infection and undergone
chest CT within 24 hours of confirmation. By quantitative CT analysis,
and collecting clinical data, we explored correlations between those
variables. Our research included 253 patients, after screening by
exclusion criteria, 171 patients were included in final cohort. The
incidence of SARS-CoV-2 associated pneumonia was 74.3%. The ROC test
results showed AUCs for leukomonocyte count, and virus genes were 0.703,
0.562, 0.567, and 0.582, respectively. GGO pattern in CT was correlated
PaO2/FiO2 ratio. Multiple linear
regression results indicated GGO was associated with
PaO2/FiO2. Meanwhile, the consolidation
was correlated with PaCO2 level. Additionally,
consolidation was also associated with neutrophil–lymphocyte ratio.
Conclusion: Lymphocyte count may be a potential marker for predicting
SARS-CoV-2 pneumonia, independent of virus load. Additionally, GGO is
correlated with hypoxia, while consolidation is associated with
PaCO2 levels and inflammation, which may affect aeration
in the lungs.
Key
words: Quantitative CT analysis, SARS-CoV-2 infection, GGO,
consolidation, PaO2/FiO2 ratio,
consolidation.
Introduction
The
World Health Organization (WHO) has report (Available from:https://covid19.who.int/. ) that pneumonia caused by
severe acute respiratory syndrome coronavirus-2
(SARS-CoV-2)
has spread to 223 countries with more than 186 million confirmed cases
and more than 4 million deaths[1]. This has become a public health
issue of international concern until now. The proportion of hospitalized
patients diagnosed with severe pneumonia and infected with SARS-CoV-2
who develop ARDS, based on oxygenation criteria, ranges from 20% to
67%[2, 3].
Computed tomography (CT) imaging is often used to confirm COVID-19 and
plays a crucial role in the diagnosis and treatment assessment of
COVID-19 due to its high sensitivity [4, 5]. The main
characteristics of COVID-19 observed through CT imaging include
bilateral pulmonary ground-glass opacity (GGO), a ”crazy paving”
pattern, airway changes, and a reversed halo sign. Other manifestations
may include consolidation and interlobular septal thickening. GGO is the
earliest and most prominent pulmonary abnormality observed, while
consolidation appears in the later stages [6,
7].
Furthermore, several studies have indicated that CT features are
associated with the level of plasma cytokines [8, 9], and even with
the prognosis of pneumonia associated with
SARS-CoV-2[6,
10]. Based on these results, we want to make it clear whether a heavy
virus load could result in significant alteration in CT manifestation
and quantitative CT analysis of different CT patterns can provide more
information about patients to help clinicians make better clinical
interventions. However, there are a few questions that need to be
answered. First, there is limited research available that links the
physiological, laboratory [such as
arterial
blood gas (ABG), complete blood count (CBC), etc.], and imaging
features of patients with
SARS-CoV-2[2].
Second, the understanding of CT imaging reports is subjective and relies
on clinical experience. Therefore, it is necessary to clarify the
relationship between CT patterns and pathophysiological states.
The
purpose of this study is to investigate the results of quantitative CT
analysis for different CT patterns in SARS-CoV-2 infected patients, and
how these relate to viral load and pathophysiological states.
Method
Study population
The local ethics board has approved this cross-sectional clinical cohort
study. Patients infected with
SARS-CoV-2
who were in charge in Qingdao Hospital, University of Health, and
Rehabilitation Sciences (Qingdao Municipal Hospital) from Jun.
9th to 15th, 2023 (the initial week
of
deblocking
for isolation in Qingdao region, China) were eligible to participate in
this research.
SARS-CoV-2
was detected by polymerase chain reaction (PCR) with a sample of throat
swab or sputum specimen and the Cycle threshold (Ct) value below 40 was
regarded as positive. Pneumonia was diagnosed according to the American
Thoracic Society guidelines for community-acquired pneumonia[11].
The inclusion criteria was adult patients, who had a confirmed
SARS-CoV-2 infection and undergone chest CT within 24 hours of
admission. Patients with
a.)
history of pulmonary surgery; b.) thoracocyllosis or rib fracture; c.)
pneumothorax with/without drainage; d.) Massive pleural effusion can’t
be analysis by CT; e.) clinical diagnosed pneumonia without PCR
validation; f.) clinical data missing were excluded. According to CT
manifestation, patients were divided into SARS-CoV-2 infection without
pneumonia and pneumonia associated with SARS-CoV-2. Clinical data were
extracted from the hospital information system (HIS), and CT raw data
(The digital imaging and communications in medicine, DICOM files) for
the final target patients were downloaded from the picture archiving and
communication system (PACS) in Qingdao Hospital, University of Health,
and Rehabilitation Sciences.
Clinical
laboratory data
The medical records were reviewed to record the patient’s medical
history and interventions upon admission. The virus load was detected by
the PCR Ct value of swab or sputum specimens for each patient. To
describe the oxygenation situation and inflammatory response of these
patients, results for ABG, fraction of inspired oxygen
(FiO2), method of oxygen inhalation, and CBC were
recorded. The ratio of partial pressure of arterial oxygen to fractional
concentration of oxygen in inspired air
(PaO2/FiO2 ratio) was also calculated.
All laboratory records were collected within a 12-hour window around the
date of the CT test.
CT
examination protocol
CT examinations were performed using a multidetector scanner without the
use of contrast medium for enhancement. Image raw data were collected by
the Somatom Sensation 64 scanner (SIEMENS, Germany). The tube voltage
used was 120 KV and the tube current was set to automatic. The pitch was
1.375 and the slice thickness was 5mm. All cases used the classic
filtered back projection method with a soft tissue kernel of B20 and a
lung kernel of B60. In all patients, a spiral acquisition was obtained
from the apex to the bottom of the lungs at the stage of end inspiratory
hold breathing. Coronal and sagittal multiplanar reconstructions were
also available in all cases. The raw data were concluded into a 250-330
mm field of view and a 512×512 reconstruction matrix. The data were then
downloaded, saved as DICOM files, and made available for analysis.
Imaging
analysis
The CT scan data were analyzed by 3D slicer software[12] (version
5.0.3, http://www.slicer.org/ ) to estimate the affected lung area
(GGO + consolidation + other patterns). The DICOM files were imported
into a dedicated medical imaging software that includes a semi-automated
segmentation algorithm for lung segmentation. Two senior doctors
reviewed the results independently, without access to patients’ clinical
information such as therapy or primary disease. In case of
discrepancies, consensus was reached after consulting to another senior
radiologist, and the results were saved as regions of interest (ROI) for
quantitative analysis. Chest CT scans measured the density of the
pulmonary voxel within ROI in Hounsfield units (HUs). Depending on the
range of HU, ROI was divided into ground-glass opacity (GGO) regions
(-500HU to -100HU), consolidations (-100HU to 100HU), and others
(aerated lung: <-500HU, etc. ). Volumetric analysis
and/or visualization in 3D Slicer via the Lung CT Analyzer project
(https://github.com/rbumm/SlicerLungCTAnalyzer/ ) were performed.
The total lung and affected volumes were estimated by reconstructing
each marked slice using Lung CT Analyzer, a plug-in integrated into 3D
Slicer[13]. (Data processing procedure was shown in Fig. 1).