Actively Recruiting
Artificial Intelligence Based System for Assessing Suspected Viral Pneumonia Related Lung Changes Using Visual Pulmonary Lesion Grading System (CT 0-4): Retrospective Study
Led by Research and Practical Clinical Center for Diagnostics and Telemedicine Technologies of the Moscow Health Care Department · Updated on 2024-07-22
563
Participants Needed
1
Research Sites
26 weeks
Total Duration
On this page
Sponsors
R
Research and Practical Clinical Center for Diagnostics and Telemedicine Technologies of the Moscow Health Care Department
Lead Sponsor
S
Sciberia Co. Ltd
Collaborating Sponsor
AI-Summary
What this Trial Is About
Researchers are evaluating an AI-based system designed to analyze chest computed tomography (CT) images for signs of interstitial changes associated with viral pneumonia, including COVID-19. This retrospective study aims to clinically validate the AI system by measuring its sensitivity, specificity, accuracy, and area under the ROC curve, comparing these metrics to those declared by the manufacturer. The goal is to confirm that the AI system's diagnostic accuracy differs by no more than 8% from the declared values. The study involves collecting a verified and labeled dataset of chest CT images, including those without pneumonia and images representing five levels of lung involvement (CT-0 to CT-4) graded by the extent of pulmonary changes. The AI software processes these images to detect pathological patterns and quantify lung damage. Expert radiologists independently assess the same images, and the AI system's performance is then evaluated against these expert assessments to determine clinical efficacy. Participants in this retrospective study are patients aged 18 and older who have undergone chest CT scans without contrast enhancement following a standardized protocol. Researchers will assess the quality and characteristics of the CT images, ensuring they meet study criteria. The primary outcomes measured up to one year after study completion include accuracy, sensitivity, specificity, and area under the ROC curve of the AI system. Additionally, the approximate volume of affected lung tissue is evaluated to understand the extent of lung involvement.
CONDITIONS
Brief Title
AI-based System for Assessing Suspected Viral Pneumonia Related Lung Changes
Who Can Participate
Eligibility Criteria
You may qualify if you...
- Patients over 18 years old
- Patients who underwent chest CT scans without contrast enhancement
- Chest CT scans performed with standardized protocol: 120 kilovolts, slice thickness max. 2 mm, rigid "lung" filter reconstruction
- CT scans of acceptable quality, performed with breath-holding and without technical or motion artifacts
- CT images containing DICOM tags for patient orientation, scan size, and image parameters
- Lung changes predominantly bilateral, located in basal and subpleural lung parts, possibly peribronchial
- For Normal group: patients without COVID-19-related CT patterns
- For Mild, Moderate, Severe, and Critical groups: patients with COVID-19-related CT patterns including ground glass opacities, pulmonary consolidation, cobblestone infiltration, hydrothorax, or combinations thereof
You will not qualify if you...
- Studies containing images with unreported CT patterns
- Examinations not conforming to DICOM format
- Examinations missing lung region imaging
- Studies with technical artifacts due to scanner malfunctions or features
- Examinations with improper patient positioning
- Studies missing DICOM tags related to scan size and image parameters
- Examinations containing metal artifacts on body or clothing
- Presence of other lung pathologies such as neoplastic disease, tuberculosis, or bacterial pneumonia
- Patients under 18 years old
AI-Screening
AI-Powered Screening
Complete this quick 3-step screening to check your eligibility
Your Study Journey
Duration - 2 to 4 weeks
Participants are screened for eligibility to participate in the trial.
1 visit (in-person)
Duration - Up to 1 year
Participants' chest CT images are retrospectively analyzed using an AI-based system to identify and quantify suspected viral pneumonia related lung changes.
No visits; retrospective image analysis
Duration - Up to 1 year
Participants' data are monitored through the evaluation of AI system performance metrics including accuracy, sensitivity, specificity, and area under the ROC curve.
No visits; ongoing data assessment
Trial Site Locations
Total: 1 location
1
Research and Practical Clinical Center for Diagnostics and Telemedicine Technologies of the Moscow Health Care Department
Moscow, Russia, 127051
Actively Recruiting
Research Team
V
Victoria Zinchenko
A
Anton Vladzymyrskyy
How is the study designed?
Study Type
OBSERVATIONAL
Masking
N/A
Allocation
N/A
Model
N/A
Primary Purpose
N/A
Number of Arms
5
Similar Trials
Frequently Asked Questions
Have more questions? Get in touch with our team for quick support
Not the Right Trial for You?
Explore thousands of other clinical trials that might be a better match.
Sign up to get personalized trial recommendations delivered to your inbox.
Already have an account? Log in here