This repository contains the C++ code of a real-time application for echocardiographic guidance. The application is built on the top of the FAST framework. Ultrasound data is streamed to the application.
The application provides intuitive feedback on how to move the ultrasound to reach standardized and reproducible apical standard views.
The objective is to place the blue/purple ultrasound sector within the red target box in the upper right figure. When the sector gets into the box, the box becomes green, and the view is standardized. A demonstration of how to use the application can be seen in this video.
Further information on how to use the application can be found in the user manual.
Clinical scanner using FAST OpenIGTLink streamer
OpenIGTLink is an open source library for real-time image streaming. It allows you to importing images from various ultrasound scanners to this application.
We have used OpenIGTLink to stream images in real-time from a GE E95 ultrasound scanner in our multiple clinical studies. The use of the application with a GE E95 scanner is demonstrated here.
Clarius scanner using the FAST Clarius streamer
Ultrasound data can be streamed directly from a Clarius ultrasound scanner using the Clarius API.
DICOM Multiframe file using the FAST DICOM Multiframe Streamer
The application can be run retrospectively on DICOM Multiframe files. A detailed procedure on how to achieve this is described here.
Image file using the FAST Image File Streamer
The application can be run retrospectively on .png and .mhd files. A detailed procedure on how to achieve this is described here.
Please see the build instructions.
This repository uses different licenses for code and model artifacts.
The code in this repository (within source) is distributed under PolyForm Noncommercial License 1.0.0. Please refer to the license terms available here or in the LICENSE_CODE.md included in the repository.
The models (within models) are distributed as open source under the Creative Commons Attribution-NonCommercial 4.0 International Public License. Please refer to the license terms available here or in the LICENSE_MODELS.txt included in the repository.
If you are using this repository, please cite the following articles:
The initial research paper on technology development:
David Pasdeloup, Sindre H. Olaisen, Andreas Østvik, Sigbjorn Sabo, Håkon N. Pettersen, Espen Holte, Bjørnar Grenne, Stian B. Stølen, Erik Smistad, Svein Arne Aase, Håvard Dalen, Lasse Løvstakken, Real-Time Echocardiography Guidance for Optimized Apical Standard Views, Ultrasound in Medicine & Biology, Volume 49, Issue 1, 2023, Pages 333-346, ISSN 0301-5629, https://doi.org/10.1016/j.ultrasmedbio.2022.09.006
The clinical validation of the technical method:
Sigbjorn Sabo, David Pasdeloup, Hakon Neergaard Pettersen, Erik Smistad, Andreas Østvik, Sindre Hellum Olaisen, Stian Bergseng Stølen, Bjørnar Leangen Grenne, Espen Holte, Lasse Lovstakken, Havard Dalen, Real-time guidance by deep learning of experienced operators to improve the standardization of echocardiographic acquisitions, European Heart Journal - Imaging Methods and Practice, Volume 1, Issue 2, September 2023, qyad040, https://doi.org/10.1093/ehjimp/qyad040
The FAST framework, on which this application is built:
E. Smistad, A. Østvik and A. Pedersen, "High Performance Neural Network Inference, Streaming, and Visualization of Medical Images Using FAST," in IEEE Access, vol. 7, pp. 136310-136321, 2019, doi: https://doi.org/10.1109/ACCESS.2019.2942441.
David Pasdeloup, Lasse Løvstakken, Andreas Østvik, Erik Smistad.