Observer performance evaluation of the feasibility of a deep learning model to detect cardiomegaly on chest radiographs.

Pranav AjmeraAmit KharatTanveer GupteRicha PantViraj KulkarniVinay A DuddalwarPurnachandra Lamghare

Published in: Acta radiologica open (2022)

Our segmentation-based AI model demonstrated high specificity (>99%) and sensitivity (80%) for CTR calculation. The performance of the radiologist on the observer performance test improved significantly with provision of AI assistance. A DL-based segmentation model for rapid quantification of CTR can therefore have significant potential to be used in clinical workflows by reducing radiologists' burden and alerting to an abnormal enlarged heart early on.

Keyphrases

deep learning
artificial intelligence
convolutional neural network
heart failure
machine learning
palliative care
atrial fibrillation
risk factors
climate change
human health
structural basis