Imaging lymphatic system using high frame rate contrast enhanced and super-resolution ultrasound : in vitro, in vivo and clinical study

The lymphatic system acts as an avenue through which cancer cells can spread from tumour. Lymph nodes (LNs) linking the lymphatic vessels play a critical role in filtering the cancer cells. Thus, the presence of metastases in LN is an important indicator for cancer staging [1]. As angiogenesis is a key feature of cancer, any variations in LN microcirculation can be a biomarker for metastasis. However, a non-invasive technique capable of visualising and quantifying vasculature and blood flow in LNs is lacking and highly desirable. Contrast-enhanced ultrasound (CEUS) utilising microbubbles shows great potential for visualising lymphatic vessels and identifying sentinel LNs. It allows for the diagnosis of metastatic LNs through the visualisation of perfusion pattern. However, current CEUS imaging techniques have several limitations that prevent the accurate identification, visualisation and diagnosisof LNs: (i) Tissue artefacts and bubble disruption can reduce the image contrast. (ii) Limited spatial and temporal resolution diminishes the amount of information that can be captured by CEUS. (iii) The slow flow in both the lymphatic system and the micro-vessels inside LNs makes Doppler-based approaches less effective. (iv) The identification of enhancement pattern is subjective and can vary between different observers. In this thesis, we aim to improve the capability of CEUS in non-invasive identification and evaluation of LNs through the use of high frame rate (HFR) imaging and super-resolution (SR) ultrasound (US) imaging. Firstly, we evaluate the feasibility of using HFR CEUS for the detection of lymphatic vessels where flow is slow. The study is carried out on a lymphatic vessel phantom. Specifically, the work investigates how CEUS lymphatic imaging is affected by several key factors such as ultrasound pressure, flow velocity probe motion and image contrast. Experiments were also conducted to observe microbubble behaviour under HFR CEUS. Our results show that (i) HFR imaging and singular value decomposition (SVD) filtering can significantly reduce tissue artefacts in the phantom at high clinical frequencies; (ii) the slow flow rate within the lymphatic system makes image contrast and signal persistence more susceptible to changes in ultrasound amplitude or mechanical index (MI), and a suitable MI value can be chosen to reach a compromise between image contrast and bubble disruption under slow flow condition; (iii) probe motion significantly decreases image contrast of the vessel, which can be improved by applying motion correction before SVD filtering; (iv) the optical observation of the impact of ultrasound pressure on HFR CEUS further confirms the importance of optimising ultrasound amplitude. Secondly, the feasibility of in vivo three-dimensional SR US imaging for the visualisation and quantification of LN microvascular and blood flow was studied in a rabbit model. In vivo studies were carried out to image popliteal LNs of two healthy male New Zealand white rabbits aged 6–8 weeks. Three-dimensional, high-frame-rate, contrast enhanced US was achieved by mechanically scanning a linear imaging probe. Individual microbubbles were identified, localised, and tracked to form three-dimensional SR images and super-resolved velocity maps. Acoustic sub-aperture processing was used to improve image contrast and to generate enhanced power Doppler and colour Doppler images. Vessel size and blood flow velocity distributions were evaluated and assessed by using paired Student’s t-test. SR images is able to reveal microvessels in the rabbit LN with branches clearly resolved up to 30 μm. This is less than half the acoustic wavelength and not resolvable by power or colour Doppler. The apparent size distribution of most vessels in the SR images was below 80 μm which agrees with micro-CT data. In contrast, most vessels detected with Doppler techniques were larger than 80 μm. The blood flow velocity distribution indicated that most of the blood flow in rabbit popliteal LN was at velocities lower than 5 mm/sec. We demonstrate that three-dimensional super-resolution US imaging using microbubbles allows non-invasive, non-ionising visualisation and quantification of LN microvascular structures and blood flow dynamics with resolution below the wave diffraction limit. This technology has great potential for further study of the physiological functions of the lymphatic system and the clinical detection of LN metastasis. Finally, we explore the potential of using SR US for distinguishing metastatic LNs from reactive LNs in human. In this study, SR US images and super resolved velocity maps of LNs are generated from patient data acquired with a clinical ultrasound system. We discuss the challenges of acquiring suitable datasets in the clinical settings and suggest solutions to overcome these challenges. For example, destruction pulses are utilised to achieve suitable microbubble concentration for super-resolution ultrasound imaging. In addition, motion correction is applied before the detection of microbubble to compensate for the significant movement in real clinical settings. After obtaining the super resolution ultrasound images, multiple morphological and functional measures are derived from the super resolution images and super resolved velocity maps. These measures were compared for the reactive and metastatic LNs and assessed using a two sample student’s t-test. Our initial results indicate that metastatic LNs have significantly higher variance in blood flow direction compared to that of reactive LNs. Different factors affecting the quantifications of super-resolution ultrasound imaging are also studied and potential solutions are provided to take these factors into account. This study demonstrates the possibility of using quantifications obtained from super resolution ultrasound images to discriminate between metastatic and reactive LNs.