Research Associate in Deep Learning Ultrasound Methods

This is an exciting opportunity for an enthusiastic researcher with knowledge in deep learning and image analysis to join an international project funded by the Wellcome Trust Innovations Flagship programme.

Ultrasound imaging is a safe, portable and comprehensive tool for monitoring and management of many different types of patients, including patients in the ICU. However, ultrasound requires, on the one hand, manual dexterity and expertise to operate the probe and find the clinically relevant views; and on the other hand, extensive training and skills to accurately interpret and analyse the resulting images. Recently, in the context of a fetal screening project (IFIND), we have integrated a number of computational methods into an engineered clinical setting that allows to connect an ultrasound system to a number of sensors and a computer to produce a computer-assisted system to help the sonographer with ultrasound examinations.

In this project, we want to extend this know-how to Intensive Care Unit in a resource-limited setting, to assist local clinicians with cardiac, lung, eye and muscle ultrasound imaging for patient management. The aims are 1) to investigate a clinically usable set-up which allows to connect a standard 2D ultrasound system to a computer for real-time analysis of ultrasound images, that remains mobile within the unit and user-friendly, 2) investigate, develop and integrate computational methods to process, in real-time and off-line, ultrasound images to find clinically relevant imaging planes and extract biometrics and other quantities from them.