Key result
A ResNet50 deep learning model using standard apical 4-chamber echocardiography views accurately differentiated constrictive pericarditis from cardiac amyloidosis with an AUC of 0.97.
Why the study?
Diagnosing constrictive pericarditis with echocardiography remains challenging for clinicians, leading researchers to evaluate a deep learning model to differentiate it from restrictive cardiomyopathy.
Does an echocardiography-based deep learning model accurately differentiate constrictive pericarditis from cardiac amyloidosis using standard apical 4-chamber views?
Observational (n=381)
No
Does an echocardiography-based deep learning model accurately differentiate constrictive pericarditis from cardiac amyloidosis using standard apical 4-chamber views?
A deep learning model using standard apical 4-chamber echocardiography views can accurately differentiate constrictive pericarditis from cardiac amyloidosis, potentially improving diagnostic workflow.
May support DL differentiation of constrictive pericarditis from cardiac amyloidosis on echocardiography; hypothesis-generating pending prospective validation.
Background Constrictive pericarditis (CP) is an uncommon but reversible cause of diastolic heart failure if appropriately identified and treated. Although echocardiography can detect CP based on characteristic cardiac motion and Doppler findings, its diagnosis remains a challenge for clinicians. Artificial intelligence (AI) may enhance identification of CP. We proposed a deep learning approach based on transthoracic echocardiography (TTE) to differentiate CP from restrictive cardiomyopathy (RCM). Methods Patients with a confirmed diagnosis of CP and cardiac amyloidosis (CA, as the representative disease of RCM) at Mayo Clinic Rochester from 1/2003-12/2021 were identified to extract baseline demographics and the apical 4 chamber (A4C) view from TTE studies. The cases were split into a 60:20:20 ratio for training, validation, and held-out test sets of the ResNet50 deep learning model. The model performance (differentiating CP and CA) was evaluated in the test set with the area under the curve (AUC). GradCAM was used for model interpretation. Results A total of 381 patients were identified, including 184 (48.3%) CP, and 197 (51.7%) CA cases. The mean age was 68.7±11.4, and 72.8% were male. ResNet50 had a performance with an AUC to differentiate the 2-class classification task (CP vs. CA, AUC 0.97). The GradCAM heatmap showed activation around the ventricular septal area. Conclusion With a standard A4C view, our AI model provides a platform for the early and accurate detection of CP, allowing for improved workflow efficiency and prompt referral for more advanced evaluation and intervention of CP.
No takes yet. Share an insight, caveat, or question.
Chao et al. (2022) conducted an observational in Constrictive Pericarditis and Cardiac Amyloidosis (n=381). Echocardiography-based deep learning model (ResNet50) was evaluated on Differentiation of constrictive pericarditis from cardiac amyloidosis (Area Under the Curve) (AUC 0.97). A ResNet50 deep learning model using standard apical 4-chamber echocardiography views accurately differentiated constrictive pericarditis from cardiac amyloidosis with an AUC of 0.97.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: