Echocardiographic Image Classification Using Deep Learning

  • Salvador Gomes Neto Universidade do Vale do Rio do Sinos
  • Samuel Armbrust Freitas Universidade do Vale do Rio do Sinos
  • Gabriel de Oliveira Ramos Universidade do Vale do Rio do Sinos


Left ventricular ejection fraction is a physiological measure obtained by evaluating the cardiac phases of systole or diastole. This parameter represents the contractile capacity of the cardiac ventricular chambers, which several methods can measure, echocardiography being the most cost-effective. The correct ejection fraction assessment is critical for diagnosing and treating most cardiovascular diseases. Although using deep learning to estimate the ejection fraction significantly improves the method’s accuracy, there are still difficulties with its extensive application for several reasons. This paper proposes a deep learning pipeline for classifying echocardiographic images in systole or diastole, comparing its performance to the state-of-the-art. The proposed pipeline features a set of pre-processing methods suitable to echocardiographic images and a convolutional neural network tuned for the considered classification task. We also introduce a novel dataset of echocardiographic images without excessive pre-selection of images, thus presenting real-life conditions. We performed several experiments to assess the performance of our approach, through which it was possible to obtain an accuracy of 97.69% and a cross-entropy loss of 0.1883. Our convolutional neural network was able to classify systolic and diastolic images with accuracy similar to the benchmark in the literature. The proposed pipelines present pre-processing methods suitable for echocardiographic images, a convolutional neural network adjusted for the considered classification. However, our network is simpler than the reference, and the dataset is closer to real-life conditions, avoiding excessive pre-selection of images.

Palavras-chave: Deep Learning, Echocardiography, Systole, Diastole, Convolutional Neural Network, Image Classification


Caroline Petitjean, J.N.D.: A review of segmentation methods in short axis cardiac mr images. Medical Image Analysis 15, 169–184 (2011)

Chang, A., Cadaret, L.M., Liu, K.: Machine learning in electrocardiography and echocardiography: technological advances in clinical cardiology. Current Cardiology Reports 22, 1–7 (2020)

Chu WK, R.D.: Fourier analysis of the echocardiogram. Phys Med Biol 23(1), 100–105 (1978)

Jeffrey Zhang, Sravani Gajjala, P.A.: Fully automated echocardiogram interpretation in clinical practice. Circulation 138, 1623–1635 (2018)

Libby, P., Bonow, R.O., Mann, D.L., Tomaselli, G.F., Bhatt, D., Solomon, S.D., Braunwald, E.: Braunwalds Heart Disease. A textbook of Cardiovascular Medicine. Elsevier, 12th edn. (2022)

Litjens, G., Ciompi, F., Wolterink, J.M., de Vos, B.D., Leiner, T., Teuwen, J., Isgum, I.: State-of-the-art deep learning in cardiovascular image analysis. JACC: Cardiovascular Imaging 12(8), 1549–1565 (2019)

Madani, A., Arnaout, R., Mofrad, M., Arnaout, R.: Fast and accurate view classification of echocardiograms using deep learning. npj Digital Medicine 1(6) (2018).

Madani, A., Arnaout, R., Mofrad, M., Arnaout, R.: Current challenges and recent updates in artificial intelligence and echocardiography. Current Cardiovascular Imaging Reports 13(5) (2020).

Ostvik, A., Smistad, E., Aase, S.A., Haugen, B.O., Lovstakken, L.: Real-time standard view classification in transthoracic echocardiography using convolutional neural networks. Ultrasound in Med. & Biol pp. 1–11 (2018).

Paaladinesh Thavendiranathan, Shizhen Liu, D.V.: Real-time full-volume 3d transthoracic echocardiography to measure lv volumes and systolic function. JACC: Cardiovascular Imaging 5(3), 239–251 (2012)

Sachpekidis, V., Papadopoulou, S.L., Kantartzi, V., Styliadis, I., Nihoyannopoulos, P.: Use of artificial intelligence for real-time automatic quantification of left ventricular ejection fraction by a novel handheld ultrasound device. European Heart Journal-Cardiovascular Imaging 23(Supplement 1), jeab289–005 (2022)

Salte, I., Oestvik, A., Smistad, E., Melichova, D., Nguyen, T., Brunvand, H., Edvardsen, T., Loevstakken, L., Grenne, B.: 545 deep learning/artificial intelligence for automatic measurement of global longitudinal strain by echocardiography. European Heart Journal-Cardiovascular Imaging 21(Supplement 1), jez319–279 (2020)

Samtani, R., Bienstock, S., Lai, A.C., Liao, S., Baber, U., Croft, L., Stern, E., Beerkens, F., Ting, P., Goldman, M.E.: Assessment and validation of a novel fast fully automated artificial intelligence left ventricular ejection fraction quantification software. Echocardiography 39(3), 473–482 (2022)

de Siqueira, V.S., Borges, M.M., Furtado, R.G., Dourado, C.N., da Costa, R.M.: Artificial intelligence applied to support medical decisions for the automatic analysis of echocardiogram images: A systematic review. Artificial intelligence in medicine 120, 102165 (2021)

WF. Armstrong, T.R.: Feigenbaum’s Echocardiography. Lippincott Williams & Wilkins, 8th edn. (2019)

Zhang, J., Gajjala, S., Agrawal, P., Tison, G.H., Hallock, L.A., Beussink-Nelson, L., Lassen, M.H., Fan, E., Aras, M.A., Jordan, C., et al.: Fully automated echocardiogram interpretation in clinical practice: feasibility and diagnostic accuracy. Circulation 138(16), 1623–1635 (2018)
GOMES NETO, Salvador; FREITAS, Samuel Armbrust; RAMOS, Gabriel de Oliveira. Echocardiographic Image Classification Using Deep Learning. In: ENCONTRO NACIONAL DE INTELIGÊNCIA ARTIFICIAL E COMPUTACIONAL (ENIAC), 20. , 2023, Belo Horizonte/MG. Anais [...]. Porto Alegre: Sociedade Brasileira de Computação, 2023 . p. 473-485. ISSN 2763-9061. DOI: