Cardiac Acquisitions for Multi-structure Ultrasound Segmentation

Alberto García-Galindo

Institute of Biomedical Research of Salamanca

Challenge overview

The goal of this project is to provide all the materials to the community to resolve the problem of echocardiographic image segmentation and volume estimation from 2D ultrasound sequences (both two and four-chamber views). To this aim, the following solutions were set-up:

• Introduction of the largest publicly-available and fully-annotated dataset for 2D echocardiographic assessment (to our knowledge). The CAMUS dataset, containing 2D apical four-chamber and two-chamber view sequences acquired from 500 patients, is made available for download.

• Deployement of a dedicated Girder online platform. This platform aims at assessing in a reproductible manner the performance of methods for the segmentation of cardiac structures (left ventricle endocardium and epicardium and left atrium borders) and the extraction of clinical indices (left ventricle volumes and ejection fraction).

More information about the project can be found here.

---

Automatic Segmentation Example

---

Encoder-Decoder Architectures

UNet (Ronnenberg, 2015)

LadderNet (Zhuang, 2019)