Development of 2D and 3D Ultrasound Super-Resolution (US-SR) Imaging for the Clinic

Lead Research Organisation: King's College London
Department Name: Imaging & Biomedical Engineering

Abstract

Cancer is among the leading causes of death worldwide. Cancers generate their own network of blood vessels to provide nutrients and oxygen to grow and spread. Detecting these developments early and treating them increases the chances of survival. However, current imaging methods are unable to detect these microscopic structures deep within the body. Therefore there is a crucial need to develop new imaging techniques that can fill this requirement. Additionally, imaging techniques which can look at the full 3D region of disease are urgently needed to reliably assess these.

The research in this proposal is designed to develop and demonstrate an ultrasound imaging technique known as ultrasound super-resolution (US-SR) in 2D and 3D in the clinic. US-SR is able to image extremely fine details of the blood vessel network, previously unseen with standard ultrasound imaging. This technique involves adding small amounts of microbubbles into the blood stream, which show up on the ultrasound images because the sound is more strongly reflected from the bubbles than other tissues. These bubbles circulate harmlessly within the vessels until they dissolve after a few minutes. By pinpointing the location of these travelling bubbles over time, we can build up an image which 'paints out' the vessel structures containing those microbubbles. The ability to see these small vessels using ultrasound, which is able to image at depth (>10 cm) in humans, has the potential to identify these important changes in the vascular network

Currently, however, US-SR has only been demonstrated in a small number of patients, it requires long scan times (in the range of 10s of minutes depending on the target) and ultrasound use in hospitals is generally limited to 2D. Within this proposed fellowship, it is my aim to firstly, develop faster ways to acquire the data needed to create these images. Secondly, to demonstrate the use of 2D US-SR in a large number of patients. And lastly, to use these developments to move 3D US-SR into the clinic.

Successful 3D clinical US-SR demonstration could propel this technique into clinical practice. Its use could provide safe, low-cost microscopic assessment of blood vessels associated with disease. This could be crucial to patients with a wide range of micro-vascular related diseases including cancer due to early diagnosis and treatment. Given that US is an affordable imaging technique compared to for example x-ray CT and MRI, this could also provide significant cost-savings for the NHS.

Technical Summary

Non-invasive imaging of the microvasculature is crucial for the early detection and intervention of diseases such as cancer and other microvascular related diseases. This proposal addresses a crucial clinical challenge: the lack of a sensitive, safe, repeatable method for detecting, characterising and monitoring such diseases. Recently developed ultrasound super-resolution (US-SR) is able to resolve microvascular details far beyond the diffraction limit at depth (>10 cm) in vivo. To transfer this technology into society, clinical translation and 3D US-SR development is urgently needed.
Aims:. This research is designed to provide advances in areas of maximal clinical relevance through systematically designed experiments, recent technological advances, and clinical guidance and support. The results will provide both validation, and a solid grounding in basic science for the fast developing field of US-SR. The core objectives are:
1)To develop novel advanced, acquisition strategies and real-time software using deep learning, advanced signal processing methods, and the activation of sparse contrast agent to provide improved detection accuracy, localisation rates, and automation.
2)To design, develop and test (in vitro and in vivo) the accuracy and clinical relevance of structural and functional parameters for disease characterisation.
3)To formalise clinical 2D US-SR and establish clinical acceptance. This involves the demonstration of the diagnostic power of US-SR parameters over large clinical datasets from existing studies.
4)To develop and implement clinical 3D US-SR with the aid of advances made in 1)-3), a healthy volunteer pilot study and existing patient studies.
By making maximal use of planned or existing trials, I will avoid the need for significant additional human involvement. Throughout this translational project, methods will be evaluated by an extensive network of clinicians and researchers, allowing continuous feedback and ongoing knowledge

Planned Impact

Patients, the NHS, clinicians, the education sector, researchers and industries will benefit from this research. The proposed research could propel ultrasound super-resolution (US-SR) into the clinical world. Clinical 2D US-SR will be implemented within the time-frame of the Fellowship. Its real-world deployment could provide repeatable, safe, low-cost microscopic assessment of vasculature. This could aid clinicians with early disease detection, diagnosis, intervention and treatment monitoring in cancer, as well as a wide range of microvascular related diseases, e.g. those associated with diabetes and ischemia. Early detection and clinical intervention can significantly increase the chances of survival for almost all cancer patients. The work has the potential therefore to enhance the quality of life for many people, and in turn contribute to the nation's improved health. Clinicians and the patient population could thus be a beneficiary of this research within the next 5-10 years. In the future, US-SR may also be able to inform a personalised treatment plan based on characterisation of the vascular network and behaviour. More sensitive, personalised diagnostic modalities would have huge worldwide benefits to the future of cancer treatment and monitoring. This could potentially occur within the next 15 years.

Given that US is an affordable technology compared to e.g. x-ray CT and MRI, this technology could also provide significant cost-savings for the NHS. In addition, real-time US-SR could aid in assessing disease conditions in a fraction of the time required for the aforementioned imaging modalities, improving clinical efficiency and reducing clinical time requirements. Furthermore, improved resolution of microvascular structure and flow may introduce new indicators to aid diagnosis and could further expand the use of contrast enhanced US in primary care. This, as well as promoting the use of 3D clinical probes, and microbubble (MB) contrast agents in both a clinical and research setting would benefit companies and industries active in US medical diagnostics and MB manufacturers, e.g. Bracco, as well as clinicians and researchers, during the time-frame of the project. This could avoid more expensive clinical examinations and hence reduce healthcare costs. Further cost-savings gained through early cancer detection and intervention would result from increased chances of successful treatment, improved survival rates, and reducing the need for costly late-stage cancer treatments. Furthermore, newly developed acquisition protocols and software offers opportunity for commercialisation and industrial development. Researchers and the education sector will also benefit during and after the project finishes through dissemination and communication about the research in the academic environment, summer schools, and through public engagement activities.

Widespread implementation of 3D US-SR will likely be slower then 2D due to the current limited availability of 3D clinical probes, however this is predicted to drastically increase in the next few years due to recent advances in ultrasound (US) hardware and software. Successful demonstration of 3D US-SR could help to expedite translation of 3D equipment to the clinic, thus benefiting both clinicians and medical diagnostic companies e.g. Phillips and GE.

Both the Fellow and Research assistant will develop investigative research skills, interpersonal and communication skills across different audiences, e.g. academics, clinicians and the public. Furthermore, skills involved in working in an interdisciplinary environment, with clinicians, patients and researchers in different areas of expertise will be hugely valuable. The fellow will also develop project management and leadership skills needed to effectively deliver this ambitious project. This will aid in providing skilled people to the workforce.

Publications

10 25 50