The study hypothesis is that low-dose computed tomography (LDCT) coupled with artificial intelligence by deep learning would generate imaging biomarkers linked to the patient's short- and medium-term prognosis. The purpose of this study is to rapidly make available an early decision-making tool (from the first hospital consultation of the patient with symptoms related to SARS-CoV-2) based on the integration of several biomarkers (clinical, biological, imaging by thoracic scanner) allowing both personalized medicine and better anticipation of the patient's evolution in terms of care organization.
Study Type
OBSERVATIONAL
Low-dose computed tomography
CHU la Timone
Marseille, France
CHU Montpellier
Montpellier, France
CHU de Nimes
Nîmes, France
CHU Poitiers
Poitiers, France
CHU Strasbourg
Vital status
Dead/alive
Time frame: Day 8
Patient requiring more than 3 liters of oxygen to maintain a saturation >95% (intensive care unit or resuscitation department)
Yes/no
Time frame: Day 8
Percentage of lung affected on CT
% ground glass and condensation calculated by deep learning
Time frame: Day 0
Percentage of lung affected by ground glass opacity on scan
% calculated by deep learning
Time frame: Day 0
Percentage of lung affected by condensation on scan
% calculated by deep learning
Time frame: Day 0
Vital status
Dead/alive
Time frame: Day 16
Vital status
Dead/alive
Time frame: Day 30
Length of hospitalization
Days
Time frame: Maximum 30 days
rehospitalization
Yes/no
Time frame: Day 30
Duration of intubation
Days
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.
Strasbourg, France
CHU Martinique
Fort-de-France, Martinique
Time frame: Day 30
Percentage of lung affected on CT
% ground glass and condensation calculated by deep learning
Time frame: Day 16
Percentage of lung affected by ground glass opacity on scan
% calculated by deep learning
Time frame: Day 16
Percentage of lung affected by condensation on scan
% calculated by deep learning
Time frame: Day 16
Software operating time
Speed of image loading and image processing depending of brand of scanner
Time frame: End of study (August 2020)
C-reactive protein levels
mg/L
Time frame: Admission Day 0
lactate dehydrogenase
U/L
Time frame: Admission Day 0
lymphocytemia
g/L
Time frame: Admission Day 0
D Dimers level
µg/L
Time frame: Admission Day 0
Time until onset of symptoms
Days
Time frame: Admission Day 0
Time between RT-PCR positive results and first scan
Hours
Time frame: Admission Day 0
Age
Years
Time frame: Admission Day 0
BMI> 30
Yes/no:
Time frame: Admission Day 0
Medical history of cardiovascular disease
Yes/no: hypertension, coronary artery disease, congestive heart failure, cardiac arrhythmia
Time frame: Admission Day 0
Diabetes
Yes/no
Time frame: Admission Day 0
Medical history of respiratory disease
Yes/no: Chronic obstructive pulmonary disease, chronic respiratory failure
Time frame: Admission Day 0
Medical history of immunosuppressed condition
Yes/no: steroid use, pre-existing immunological condition, current chemotherapy for cancer
Time frame: Admission Day 0
Current or previous history of smoking
Yes/no:
Time frame: Admission Day 0
Calculate a prognostic score from clinical, biological and CT parameters
Deep learning algorithm
Time frame: Day 8
Calculate a prognostic score from clinical and biological parameters only
Deep learning algorithm
Time frame: Day 8
Compare receiver operating curves of prognostic scores with and without CT parameters
Time frame: Day 8