Preoperative MR imaging in endometrial cancer patients provides valuable information on local tumor extent, which routinely guides choice of surgical procedure and adjuvant therapy. Furthermore, whole-volume tumor analyses of MR images may provide radiomic tumor signatures potentially relevant for better individualization and optimization of treatment. We apply a convolutional neural network for automatic tumor segmentation in endometrial cancer patients, enabling automated extraction of tumor texture parameters and tumor volume. The network was trained, validated and tested on a cohort of 139 endometrial cancer patients based on preoperative pelvic imaging. The algorithm was able to retrieve tumor volumes comparable to human expert level (likelihood-ratio test, $$p = 0.06$$p = 0.06 ). The network was also able to provide a set of segmentation masks with human agreement not different from inter-rater agreement of human experts (Wilcoxon signed rank test, $$p=0.08$$p = 0.08 , $$p=0.60$$p = 0.60 , and $$p=0.05$$p = 0.05 ). An automatic tool for tumor segmentation in endometrial cancer patients enables automated extraction of tumor volume and whole-volume tumor texture features. This approach represents a promising method for automatic radiomic tumor profiling with potential relevance for better prognostication and individualization of therapeutic strategy in endometrial cancer.

A large variety of severe medical conditions involve alterations in microvascular circulation. Hence, measurements or simulation of circulation and perfusion has considerable clinical value and can be used for diagnostics, evaluation of treatment efficacy, and for surgical planning. However, the accuracy of traditional tracer kinetic one-compartment models is limited due to scale dependency. As a remedy, we propose a scale invariant mathematical framework for simulating whole brain perfusion. The suggested framework is based on a segmentation of anatomical geometry down to imaging voxel resolution. Large vessels in the arterial and venous network are identified from time-of-flight (ToF) and quantitative susceptibility mapping (QSM). Macro-scale flow in the large-vessel-network is accurately modelled using the Hagen-Poiseuille equation, whereas capillary flow is treated as two-compartment porous media flow. Macro-scale flow is coupled with micro-scale flow by a spatially distributing support function in the terminal endings. Perfusion is defined as the transition of fluid from the arterial to the venous compartment. We demonstrate a whole brain simulation of tracer propagation on a realistic geometric model of the human brain, where the model comprises distinct areas of grey and white matter, as well as large vessels in the arterial and venous vascular network. Our proposed framework is an accurate and viable alternative to traditional compartment models, with high relevance for simulation of brain perfusion and also for restoration of field parameters in clinical brain perfusion applications. An accurate simulation of blood-flow in the human brain can be used for improved diagnostics and assignment of personalized treatment regimes. However, current algorithms are limited to simulation of blood flow within tumours only, and in terms of parameter estimation, traditional compartment models have limited accuracy due to lack of spatial connectivity within the models. As a remedy, we propose a data-driven computational fluid dynamics model where the geometric domains for simulation are defined from state-of-the art MR acquisitions enabling a segmentation of large arteries and veins. In the capillary tissue we apply a two-compartment porous media model, where the perfusion is pressure-driven and is defined as the transition of blood from arterial to venous side. In addition, we propose a model for dealing with the intermediate scale problem where the vessels are undetectable and the flow does not adhere to requirements of porous media flow. For this scale, we propose a support function distributing the fluid in a nearby region around the vessel terminals. Combining these elements, we have developed a novel full human brain blood-flow simulator.

Uterine cervical cancer (CC) is the most common gynecologic malignancy worldwide. Whole-volume radiomic profiling from pelvic MRI may yield prognostic markers for tailoring treatment in CC. However, radiomic profiling relies on manual tumor segmentation which is unfeasible in the clinic. We present a fully automatic method for the 3D segmentation of primary CC lesions using state-of-the-art deep learning (DL) techniques. In 131 CC patients, the primary tumor was manually segmented on T2-weighted MRI by two radiologists (R1, R2). Patients were separated into a train/validation (n = 105) and a test- (n = 26) cohort. The segmentation performance of the DL algorithm compared with R1/R2 was assessed with Dice coefficients (DSCs) and Hausdorff distances (HDs) in the test cohort. The trained DL network retrieved whole-volume tumor segmentations yielding median DSCs of 0.60 and 0.58 for DL compared with R1 (DL-R1) and R2 (DL-R2), respectively, whereas DSC for R1-R2 was 0.78. Agreement for primary tumor volumes was excellent between raters (R1-R2: intraclass correlation coefficient (ICC) = 0.93), but lower for the DL algorithm and the raters (DL-R1: ICC = 0.43; DL-R2: ICC = 0.44). The developed DL algorithm enables the automated estimation of tumor size and primary CC tumor segmentation. However, segmentation agreement between raters is better than that between DL algorithm and raters.

Dynamic MR image recordings (DCE-MRI) of moving organs using bolus injections create two different types of dynamics in the images: (i) spatial motion artifacts due to patient movements, breathing and physiological pulsations that we want to counteract and (ii) signal intensity changes during contrast agent wash-in and wash-out that we want to preserve. Proper image registration is needed to counteract the motion artifacts and for a reliable assessment of physiological parameters. In this work we present a partial differential equation-based method for deformable multimodal image registration using normalized gradients and the Fourier transform to solve the Euler-Lagrange equations in a multilevel hierarchy. This approach is particularly well suited to handle the motion challenges in DCE-MRI time series, being validated on ten DCE-MRI datasets from the moving kidney. We found that both normalized gradients and mutual information work as high-performing cost functionals for motion correction of this type of data. Furthermore, we demonstrated that normalized gradients have improved performance compared to mutual information as assessed by several performance measures. We conclude that normalized gradients can be a viable alternative to mutual information regarding registration accuracy, and with promising clinical applications to DCE-MRI recordings from moving organs.

Objective: Chronic kidney disease (CKD) is a serious medical condition characterized by gradual loss of kidney function. Early detection and diagnosis is mandatory for adequate therapy and prognostic improvement. Hence, in the current pilot study we explore the use of image registration methods for detecting renal morphologic changes in patients with CKD. Methods: Ten healthy volunteers and nine patients with presumed CKD underwent dynamic T1 weighted imaging without contrast agent. From real and simulated dynamic time series, kidney deformation fields were estimated using a poroelastic deformation model. From the deformation fields several quantitative parameters reflecting pressure gradients, and volumetric and shear deformations were computed. Eight of the patients also underwent a kidney biopsy as a gold standard. Results: We found that the absolute deformation, normalized volume changes, as well as pressure gradients correlated significantly with arteriosclerosis from biopsy assessments. Furthermore, our results indicate that current image registration methodologies are lacking sensitivity to recover mild changes in tissue stiffness. Conclusion: Image registration applied to dynamic time series correlated with structural renal changes and should be further explored as a tool for invasive measurements of arteriosclerosis. Significance: Under the assumption that the proposed framework can be further developed in terms of sensitivity and specificity, it can provide clinicians with a non-invasive tool of high spatial coverage available for characterization of arteriosclerosis and potentially other pathological changes observed in chronic kidney disease.

Background: High repeatability, accuracy, and precision for renal function measurements need to be achieved to establish renal dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) as a clinically useful diagnostic tool. Purpose: To investigate the repeatability, accuracy, and precision of DCE-MRI measured renal perfusion and glomerular filtration rate (GFR) using iohexol–GFR as the reference method. Material and Methods: Twenty healthy non-smoking volunteers underwent repeated DCE-MRI and an iohexol–GFR within a period of 10 days. Single-kidney (SK) MRI measurements of perfusion (blood flow, F_b) and filtration (GFR) were derived from parenchymal intensity time curves fitted to a two-compartment filtration model. The repeatability of the SK-MRI measurements was assessed using coefficient of variation (CV). Using iohexol–GFR as reference method, the accuracy of total MR–GFR was determined by mean difference (MD) and precision by limits of agreement (LoA). Results: SK-F_b (MR1, 345 ± 84; MR2, 371 ± 103 mL/100 mL/min) and SK–GFR (MR1, 52 ± 14; MR2, 54 ± 10 mL/min/1.73 m²) measurements achieved a repeatability (CV) in the range of 15–22%. With reference to iohexol–GFR, MR–GFR was determined with a low mean difference but high LoA (MR1, MD 1.5 mL/min/1.73 m², LoA [−42, 45]; MR2, MD 6.1 mL/min/1.73 m², LoA [−26, 38]). Eighty percent and 90% of MR–GFR measurements were determined within ± 30% of the iohexol–GFR for MR1 and MR2, respectively. Conclusion: Good repeatability of SK-MRI measurements and good agreement between MR–GFR and iohexol–GFR provide a high clinical potential of DCE-MRI for renal function assessment. A moderate precision in MR-derived estimates indicates that the method cannot yet be used in clinical routine.