Diagnosing cardiac amyloidosis (CA) on echocardiography can be challenging due to the imaging overlap between CA and more prevalent causes of a hypertrophic phenotype. This study sought to:

(1) evaluate the performance of artificial-intelligence (AI) derived measurements incorporated into the established multiparametric echocardiographic scoring system to detect CA;

(2) develop and validate an AI-based deep-learning model for video-based detection of CA on echocardiography.

Key Findings:

  • The study is one of the largest and most internationally diverse CA echocardiography studies to date, comprising 5,776 patients (2,756 CA, 3,020 controls) across training and external test cohorts from the UK, USA, Taiwan, and Japan.
  • The fully automated deep-learning model achieved AUC 0.93 (95% CI, 0.91-0.95) in the US cohort and outperformed the AI-derived multiparametric score (AUC 0.88; 95% CI, 0.85-0.90), a difference that was statistically significant (P < 0.001).
  • In the US (Duke) cohort, the deep-learning model reached diagnostic accuracy of 83.2%, sensitivity of 81.7%, and specificity of 84.0%; in the Japan (NCVC) cohort, accuracy was 85.1% with comparable sensitivity and specificity.
  • The multiparametric score using AI-derived measurements also performed well, achieving accuracy of 79.5% (sensitivity 75.4%, specificity 81.5%) in the US cohort and 79.7% (sensitivity 81.6%, specificity 78.1%) in Japan, demonstrating consistent performance across globally diverse populations.
  • The deep-learning model classified a greater proportion of patients than the multiparametric score, making it more applicable in routine clinical practice where complete parameter sets may not always be available.

 

 

Read the full publication →

 

Ioannou, A., Khouri, M. G., Kitai, T., Vemulapalli, S., Hung, C.-L., Lim, S. C., Frost, M., Tee, W. W., Mansell, J., Sheikh, A., Venneri, L., Razvi, Y., Porcari, A., Martinez-Naharro, A., Rauf, M. U., Lachmann, H., Hawkins, P. N., Wechelakar, A., Moody, W., & Bandera, F. (2026). Diagnosis of Cardiac Amyloidosis on Echocardiography Using Artificial Intelligence. Circulation: Cardiovascular Imaging. https://doi.org/10.1161/circimaging.125.018991