Back to the list
Congress: ECR24
Poster Number: C-17087
Type: EPOS Radiologist (scientific)
DOI: 10.26044/ecr2024/C-17087
Authorblock: F. Souschek, P. Mildenberger, L. Müller, T. Jorg, M. C. Halfmann; Mainz/DE
Disclosures:
Fabio Souschek: Nothing to disclose
Peter Mildenberger: Nothing to disclose
Lukas Müller: Nothing to disclose
Tobias Jorg: Nothing to disclose
Moritz Christian Halfmann: Nothing to disclose
Keywords: Artificial Intelligence, Thorax, Plain radiographic studies, Comparative studies, Quality assurance
Results

Despite generally lower diagnostic performance in supine chest radiographs compared to PA radiographs, sensitivity and specificity were moderate to high in both groups for detecting lung opacities (85% vs. 92%; 63% vs. 95%), consolidation (80% vs. 75%; 81% vs. 97%), and pleural effusion (77% vs. 87%; 82% vs. 96%, respectively).  Lower diagnostic accuracy was observed in detecting pulmonary edema (59% vs. 25%; 61% vs. 100%), pneumothorax (0 % vs. 20%; 100% vs. 100%), and cardiomegaly (21% vs. 46%; 75% vs. 100%).

Fig 2: Sensitivity of AI tool
Fig 3: Specificity of AI tool

Positive predictive values were moderate to high for PA radiographs (60-100%) and markedly lower for supine radiographs (0-88%), even though the prevalence of positive target findings was considerably lower for PA radiographs.

Fig 4: PPV of AI tool
Negative predictive values were high (91-99%) for all target findings in PA radiographs. However, values were lower in supine radiographs (38-86%), with cardiomegaly having the lowest negative predictive value (38%) and consolidation the highest (86%). The lower prevalence of positive target findings in PA radiographs partially explains the lower NPV for supine radiographs.
Fig 5: NPV of AI tool
 
GALLERY