Benjamin Irving

Pieces-of-parts for supervoxel segmentation with global context: Application to DCE-MRI tumour delineation

Highlights • An automatic segmentation method is proposed for dynamic contrast enhanced MRI• We introduce perfusion-supervoxels to over-segment DCE-MRI volumes, and pieces-ofparts to add anatomical constraints to supervoxel segmentations• This method achieves promising results for the underexplored area of automatic rectal tumour segmentation from DCE-MRI scans.

Automated Colorectal Tumour Segmentation in DCE-MRI Using Supervoxel Neighbourhood Contrast Characteristics

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is a powerful protocol for assessing tumour progression from changes in tissue contrast enhancement. Manual colorectal tumour delineation is a challenging and time consuming task due to the complex enhancement patterns in the 4D sequence. There is a need for a consistent approach to colorectal tumour segmentation in DCE-MRI and we propose a novel method based on detection of the tumour from signal enhancement characteristics of homogeneous tumour subregions and their neighbourhoods. Our method successfully detected 20 of 23 cases with a mean Dice score of 0.68 +/- 0.15 compared to expert annotations, which is not significantly different from expert inter-rater variability of 0.73 +/- 0.13 and 0.77 +/- 0.10. In comparison, a standard DCE-MRI tumour segmentation technique, fuzzy c-means, obtained a Dice score of 0.28 +/- 0.17.

Comparison of linear and nonlinear implementation of the compartmental tissue uptake model for dynamic contrast‐enhanced MRI

Fitting tracer kinetic models using linear methods is much faster than using their nonlinear counterparts, although this comes often at the expense of reduced accuracy and precision. The aim of this study was to derive and compare the performance of the linear compartmental tissue uptake (CTU) model with its nonlinear version with respect to their percentage error and precision.

Myocardial scar quantification using SLIC supervoxels - parcellation based on tissue characteristic strains

Abnormal myocardial motion occurs in many cardiac pathologies, though in different ways, depending on the disease, some of which can result in negative clinical outcomes. Therefore, a better understanding of the contractile capability of the tissue is crucial in providing an improved and patient-specific clinical outcome [4]. Cardiovascular Magnetic Resonance Imaging (CMR) is considered the gold standard for the assessment of cardiac function and has the potential to also be used for routine tissue strain analysis because of its high availability in clinical practice. In this study we estimate the local strain in myocardial tissue over a cardiac cycle using cine MRI imaging to perform the analysis. To quantify the tissue displacement, we use the diffeomorphic demons registration algorithm [15] in a multi-step 3D registration, for the minimization of cumulative errors propagation. Using the displacement gradient of the deformation, individual voxel strain curves are computed. We present a novel method for parcellating the myocardium into regions based on the strain behaviour of clusters of voxels. We define the supervoxels using the Simple Linear Iterative Clustering (SLIC) algorithm [1] inside a predefined mask. The results are consistent with late gadolinium enhancement scar identification.

Segmentation of Vasculature from Fluorescently Labeled Endothelial Cells in Multi-Photon Microscopy Images

Vasculature is known to be of key biological significance, especially in the study of tumors. As such, considerable effort has been focused on the automated segmentation of vasculature in medical and pre-clinical images. The majority of vascular segmentation methods focus on bloodpool labeling methods; however, particularly, in the study of tumors, it is of particular interest to be able to visualize both the perfused and the non-perfused vasculature. Imaging vasculature by highlighting the endothelium provides a way to separate the morphology of vasculature from the potentially confounding factor of perfusion. Here, we present a method for the segmentation of tumor vasculature in 3D fluorescence microscopic images using signals from the endothelial and surrounding cells. We show that our method can provide complete and semantically meaningful segmentations of complex vasculature using a supervoxel-Markov random field approach. We show that in terms of extracting meaningful segmentations of the vasculature, our method outperforms both state-of-the-art method, specific to these data, as well as more classical vasculature segmentation methods.

Personalised Estimation of the Arterial Input Function for Improved Pharmacokinetic Modelling of Colorectal Cancer Using dceMRI

dceMRI is becoming a key modality for tumour characterisation and monitoring of response to therapy, because of the ability to identify the underlying tumour physiology. Pharmacokinetic PK models relate the contrast enhancement seen in dceMRI to physiological parameters but require accurate measurement of the AIF, the time-dependant contrast concentration in blood plasma. In this study, a novel method is introduced that overcomes the challenges of direct AIF measurement, by automatically estimating the AIF from the tumour tissue. This approach was evaluated on synthetic data 10% noise and achieved a relative error in K trans and k ep of 11.8 ±3.5% and 25.7 ±4.7 %, respectively, compared to 41 ±15 % and 60 ±32 % using a population model. The method improved the fit of the PK model to clinical colorectal cancer cases, was stable for independent regions in the tumour, and showed improved localisation of the PK parameters. This demonstrates that personalised AIF estimation can lead to more accurate PK modelling.

Improving In Vivo High-Resolution CT Imaging of the Tumour Vasculature in Xenograft Mouse Models through Reduction of Motion and Bone-Streak Artefacts

Introduction Preclinical in vivo CT is commonly used to visualise vessels at a macroscopic scale. However, it is prone to many artefacts which can degrade the quality of CT images significantly. Although some artefacts can be partially corrected for during image processing, they are best avoided during acquisition. Here, a novel imaging cradle and tumour holder was designed to maximise CT resolution. This approach was used to improve preclinical in vivo imaging of the tumour vasculature. Procedures A custom built cradle containing a tumour holder was developed and fix-mounted to the CT system gantry to avoid artefacts arising from scanner vibrations and out-of-field sample positioning. The tumour holder separated the tumour from bones along the axis of rotation of the CT scanner to avoid bone-streaking. It also kept the tumour stationary and insensitive to respiratory motion. System performance was evaluated in terms of tumour immobilisation and reduction of motion and bone artefacts. Pre- and post-contrast CT followed by sequential DCE-MRI of the tumour vasculature in xenograft transplanted mice was performed to confirm vessel patency and demonstrate the multimodal capacity of the new cradle. Vessel characteristics such as diameter, and branching were quantified. Results Image artefacts originating from bones and out-of-field sample positioning were avoided whilst those resulting from motions were reduced significantly, thereby maximising the resolution that can be achieved with CT imaging in vivo. Tumour vessels ≥ 77 μm could be resolved and blood flow to the tumour remained functional. The diameter of each tumour vessel was determined and plotted as histograms and vessel branching maps were created. Multimodal imaging using this cradle assembly was preserved and demonstrated. Conclusions The presented imaging workflow minimised image artefacts arising from scanner induced vibrations, respiratory motion and radiopaque structures and enabled in vivo CT imaging and quantitative analysis of the tumour vasculature at higher resolution than was possible before. Moreover, it can be applied in a multimodal setting, therefore combining anatomical and dynamic information.

Functional parameters derived from magnetic resonance imaging reflect vascular morphology in preclinical tumors and in human liver metastases.

Purpose: Tumor vessels influence the growth and response of tumors to therapy. Imaging vascular changes in vivo using dynamic contrast-enhanced MRI (DCE-MRI) has shown potential to guide clinical decision making for treatment. However, quantitative MR imaging biomarkers of vascular function have not been widely adopted, partly because their relationship to structural changes in vessels remains unclear. We aimed to elucidate the relationships between vessel function and morphology in vivo. Experimental Design: Untreated preclinical tumors with different levels of vascularization were imaged sequentially using DCE-MRI and CT. Relationships between functional parameters from MR (iAUC, Ktrans, and BATfrac) and structural parameters from CT (vessel volume, radius, and tortuosity) were assessed using linear models. Tumors treated with anti-VEGFR2 antibody were then imaged to determine whether antiangiogenic therapy altered these relationships. Finally, functional–structural relationships were measured in 10 patients with liver metastases from colorectal cancer. Results: Functional parameters iAUC and Ktrans primarily reflected vessel volume in untreated preclinical tumors. The relationships varied spatially and with tumor vascularity, and were altered by antiangiogenic treatment. In human liver metastases, all three structural parameters were linearly correlated with iAUC and Ktrans. For iAUC, structural parameters also modified each others effect. Conclusions: Our findings suggest that MR imaging biomarkers of vascular function are linked to structural changes in tumor vessels and that antiangiogenic therapy can affect this link. Our work also demonstrates the feasibility of three-dimensional functional–structural validation of MR biomarkers in vivo to improve their biological interpretation and clinical utility. Clin Cancer Res; 24(19); 4694–704. ©2018 AACR.

Regional Assessment of Liver Disease Progression and Response to Therapy by Multi-time Point m-SLIC Correspondence

Liver disease has reached worryingly high levels worldwide and there is a need for better analysis to monitor progression of disease and response to therapy. Quantitative imaging such as corrected T1 and PDFF can accurately quantify levels of inflammation/fibrosis and fat. In this study we develop a method to assess regional change throughout the liver to characterise disease change. We show that this method is stable in healthy test-retest cases but is able to characterise change in disease in autoimmune hepatitis cases.

Novel Quantitative Magnetic Resonance Imaging Features with Liver Function Tests to Distinguish Parenchymal and Biliary Disease

Liver disease affects millions of people worldwide and auto-immune disease in particular has unmet needs for improvement of non-invasive methods for risk-stratification. Especially in cases where clinical markers are inconclusive. In this study we develop novel imaging features for quantitative MRI and show that these features improve the differentiation of AIH from biliary disease in challenging cases, where including imaging features with clinical markers improved the AUROC from 0.76 to 0.85.