Public Funding for Intelligent Ultrasound Limited

A well recognised imaging modality, Ultrasound (US) is mobile, rapid, gives results in real time and avoids harmful radiation, such as X-rays. It is also much cheaper than magnetic resonance imaging. As the cost of portable medical US devices has reduced significantly there is a great opportunity for it to become more widely used for rapid diagnosis, and to reduce other more costly or harmful, imaging methods. However, US requires expertise to acquire and interpret images: the greatest barrier for wider adoption world-wide of medical US is no longer the need to improve imaging quality, but the need for highly trained personnel to acquire and interpret images. Training programmes have not kept up with demand, and there is no way to provide continuous monitoring of the performance of sonographers. This project aims at making US suitable for every day use by less expert healthcare professionals, particularly in community care settings. To do this, Intelligent Ultrasound (IU) Ltd has developed automated software technology and commercial know-how to check that obstetric US scans are fit-for-purpose. The proposed technical feasibility project will develop software to guide a non-expert to interpret obstetric US video in real-time as well as an expert. This new US guidance technology - built on IU's underpinning technology platform of deep-learning based US image interpretation - will automatically guide a non-expert to a high quality diagnostic plane in real-time. Innovate UK funding will address a global need and open up a totally new market for IU: supporting US closer to home.

The diagnostic quality of ultrasound scans of the same patient can vary significantly, depending on the level of skill and expertise of the person scanning the patient and the equipment used. This variation in quality can lead to patients not being diagnosed correctly, having to be recalled for a rescan or being unnecessarily referred for more invasive and costly tests. This can make diagnosis more expensive delay treatment and make hospitals inefficient. Users of ultrasound equipment are encouraged to follow guidelines about how to get the best quality scans and clinical studies have demonstrated the benefit of this approach. However, routine reviews of scan quality are not undertaken because of the the time and cost involved, so the variations in scan quality and their potentially serious consequences persist. In a collaboration between Intelligent Ultrasound Ltd and Oxford University, the AQABUS project will develop and test an automated system ("Intelligent Quality Assurance" or "IQA") for the assessment of ultrasound scan quality. IQA will provide hospitals with a means to assess scans (that is, to determine their fitness for purpose), identify staff training needs and monitor of the impact of any training undertaken. The project will require innovations that include new software to automate the analysis and scoring of ultrasound scans, and computer-based techniques of artificial intelligence known as machine learning.