EchoNous will develop a process for automated DVT diagnosis based on wearable ultrasound. The algorithms that will be developed to achieve this goal are the following: (1) real-time image grading; (2) real-time labelling of vascular structures during acquisition; (3) compression guidance and automatic vein compressibility analysis. The aforementioned algorithms will be based on deep neural networks developed by EchoNous for their point-of-care commercial ultrasound systems, which is already optimized to run efficiently on computationally resource-limited hardware such as mobile devices.
