Fast 3D Super-Resolution Ultrasound Imaging Through Acoustic Activation and Deactivation of Nanodroplets

Lead Research Organisation: Imperial College London
Department Name: Bioengineering


The microvasculature plays a crucial role in the functioning of healthy tissue throughout the body. Tumours and many other diseases, such as diabetes and coronary heart disease, cause changes in the distribution of microvessels and/or the flow within them. Detection of subtle structural and functional changes in these vessels would thus potentially enable early detection of cancer and other diseases and increase the chances of successful treatment and survival rates. Furthermore, by detecting changes in the microvasculature during treatment of these diseases, doctors may be able to identify at an early stage patient-specific treatment strategies while monitoring responses (or the lack of response) to different drugs. Current clinical imaging modalities cannot adequately resolve these tiny vessels beyond depths of a few millimetres inside the tissue. Hence there is an urgent clinical need for a new imaging method that can provide high spatial and temporal resolution at relevant tissue depths.

Optical super-resolution has revolutionised the field of optical florescence microscopy by combining information from multiple frames to achieve a single super-resolved image and was the subject of the 2014 Nobel Prize in Chemistry. However, such optical techniques only have a limited penetration depth (<1 mm) and are therefore not suitable for imaging humans in the clinic. We have developed ultrasound super-resolution imaging with contrast agents (microbubbles) with a resolution of several times better than existing clinical ultrasound imaging (down to tens of micrometres resolution at depths of several centimetres). However this approach currently requires long ultrasound data acquisition times (minutes) as time must be allowed for sparsely distributed flowing agents to traverse the full field of view. It is also challenging to image the vasculature in 3D using this approach due to the huge amount of data generated (up to TBs per second) that poses significant hardware challenges in transferring and processing such data. These shortcomings significantly limit the clinical translation of super-resolution ultrasound.

In this project, we are proposing a new technology to enable super-resolution at imaging rates up to two orders of magnitude faster than is currently possible, thereby making it becomes suitable for clinical use. To achieve this, we will replace conventional microbubble contrast agents with new "nanodroplet" agents whose in vivo imaging signal can be switched on and off acoustically in a controlled way, removing the need to use low concentrations and wait for them to flow through the entire region of interest. Second, we will use a new transducer technology with more than ten thousand elements linked in a specific way for 3D imaging that will enable rapid data capture. We will also develop optimised and fast signal processing algorithms and codes that will enable accurate super-resolution imaging and live feedback suitable for practical use. We aim to be the first to demonstrate this fast 3D super-resolution technology in vivo.

The proposed technique promises non-invasive, safe and fast microscopic assessment of vasculature in deep tissue, which could prove highly valuable to diagnosis, prediction, and intervention in a wide range of diseases.

Planned Impact

The proposed method aims to realize fast 3D microscopic imaging in vivo in deep tissue of microvasculature on a clinically relevant timescale (up to two orders of magnitude faster than existing techniques). In the short term, this novel microvascular imaging technology will benefit pre-clinical research on cancer and vascular diseases by offering a new qualitative and quantitative imaging tool for diagnosis, assessment and monitoring of such diseases.

When the technology is proven to be useful, we will start clinical trials through follow-on grants. At this stage, the main impact will be on clinical research of a wide range of micro-vascular related diseases, such as cancer, ischaemia and peripheral vascular diseases where micro-vascular characteristics are biomarkers. Long term beneficiaries of this research will be patients benefiting from earlier diagnosis, more specific treatments, monitoring of treatment efficacy and ultimately improved outcomes. Ultrasound imaging is an affordable, non-invasive and real-time technology that could provide significant cost and time-savings for the NHS in comparison to other more expensive imaging modalities such as CT and MRI.

Our project will train three post-doctoral researchers in a broad spectrum of sought-after skills ranging from nanodroplet chemistry and fabrication, ultrasound transducer development, ultrasound transmit and receive pulse sequence optimisation, 3D image reconstruction and the development of fast super-resolution image analysis software. It will also provide experience of the development and application of quantitative image analysis approaches to assessing microvasculature. These tools and skills are highly valuable to commercial medical imaging R&D labs and will increase the productivity output and competitiveness of the UK.

We will look for commercialization opportunities in close partnership with our institutional intellectual property and licensing teams at Imperial at all stages of the project. There are opportunities to file patents in the final sonoswitchable droplets, and opportunities to commercialise software for nanodroplet localisation and super-resolution image processing. We have already established links with three leading companies in pre-clinical and clinical ultrasound imaging. Bracco is the key player in the development and application of clinical ultrasound contrast agents. Philips has been one of the pioneers in the commercialisation of clinical ultrasound scanner systems and contrast enhanced imaging. FujiFilm VisualSonics is the leading preclinical ultrasound scanner developer with a special interest in contrast enhanced non-linear imaging. We will explore opprotunities with these industry links.


10 25 50
Title ULTRA-SR: Synthetic Data 
Description ULTRA-SR: Ultrasound Localisation and TRacking Algorithms for Super Resolution The ULTRA-SR Challenge was held at the 2022 IEEE International Ultrasound Symposium in Venice, Italy. The aim of the ULTRA-SR challenge is to extensively evaluate the performance of localization and tracking algorithms for Ultrasound Localization Microscopy. Two realistic simulated datasets of microvascular flow generated, using flow physics, acoustic field simulation with low and high clinical ultrasound frequencies, integrated with non-linear bubble dynamics. The ULTRA-SR Challenge dataset is publicly available for researchers to benchmark new algorithms and software. Additional information is available at: Pre-print paper to cite which describes the data generation: 
Type Of Material Database/Collection of data 
Year Produced 2022 
Provided To Others? Yes  
Impact The data was generated and used for the first international challenge on super-resolution ultrasound localisation micriscopy, as part of the IEEE International Ultrasonics Symposium in Venice 2022 which has over 1500 attendees. The data offers a valuable dataset with ground truth information so researchers are able to benchmark their localisation and tracking algorithms and develop new algorithms. The challenge has attracted interests from over 100 researchers over the world, and has over 50 challenge entries. The data was uploaded for broader use in November 2022 and in a few months has already been downloaded over 500 times.