MITEA or (MR-Informed Three-dimensional Echocardiography Analysis) dataset consists of annotated 3D echocardiography (3DE) data using labels derived from paired CMR scans acquired in a mixed cohort of 134 human subjects (82 healthy controls and 52 patients with acquired cardiac disease). For each subject, two 3DE scans (scan and rescan, in a randomised order) are available, at 2 image frames corresponding to end-diastole and end-systole. For each image, segmentations are also provided. Labelled regions are of the left ventricular myocardium (class value 1) and cavity (class value 2), which can be used for the quantification of LV systolic function and mass. A pre-trained deep learning model (based on nnU‑Net) is also made available.

For additional dataset characteristics and preliminary model validation, researchers are referred to the original publication:

Zhao, D., Ferdian, E., Maso Talou, G. D., Gilbert, K., Quill, G. M., Wang, V. Y., Babarenda Gamage, T. P., Pedrosa, J., D’hooge, J., Sutton, T. M., Lowe, B. S., Legget, M. E., Ruygrok, P. N., Doughty, R. N., Young, A. A., Nash, M. P. (2023). MITEA: A dataset for segmentation of the left ventricle in 3D echocardiography using subject-specific labels from cardiac magnetic resonance imaging. Frontiers in Cardiovascular Medicine. Vol. 9, p. 3673.


Please submit this form to get the data. We will forward your request to the Data Contributor for the approval of using the data.


  1. Martyn Nash – University of Auckland, New Zealand
  2. Alistair Young – King’s College London, UK
  3. Debbie Zhao – University of Auckland, New Zealand