Resting state MRI reveals spontaneous physiological fluctuations in the kidney and tracks diabetic nephropathy in rats.

The kidneys maintain fluid-electrolyte balance and excrete waste in the presence of constant fluctuations in plasma volume and systemic blood pressure. The kidneys perform these functions to control capillary perfusion and glomerular filtration by modulating the mechanisms of autoregulation. An effect of these modulations are spontaneous, natural fluctuations in nephron perfusion. Numerous other mechanisms can lead to fluctuations in perfusion and flow. The ability to monitor these spontaneous physiological fluctuations in vivo could facilitate the early detection of kidney disease. The goal of this work was to investigate the use of resting- state magnetic resonance imaging (rsMRI) to detect spontaneous physiological fluctuations in the kidney. We performed rsMRI of rat kidneys in vivo over 10 minutes, applying motion correction to resolve time series in each voxel. We observed spatially variable, spontaneous fluctuations in rsMRI signal between 0-0.3 Hz, in frequency bands also associated with autoregulatory mechanisms. We further applied rsMRI to investigate changes in these fluctuations in a rat model of diabetic nephropathy. Spectral analysis was performed on time series of rsMRI signal in kidney cortex and medulla. Power from spectra in specific frequency bands from kidney cortex correlated with severity of glomerular pathology caused by diabetic nephropathy. Finally, we investigated the feasibility of using rsMRI of the human kidney in two participants, observing the presence of similar, spatially variable fluctuations. This approach may enable a range of preclinical and clinical investigations of kidney function, and facilitate the development of new therapies to improve outcomes in patients with kidney disease.

BlobCUT: A Contrastive Learning Method to Support Small Blob Detection in Medical Imaging.

Medical imaging-based biomarkers derived from small objects (e.g., cell nuclei) play a crucial role in medical applications. However, detecting and segmenting small objects (a.k.a. blobs) remains a challenging task. In this research, we propose a novel 3D small blob detector called BlobCUT. BlobCUT is an unpaired image-to-image (I2I) translation model that falls under the Contrastive Unpaired Translation paradigm. It employs a blob synthesis module to generate synthetic 3D blobs with corresponding masks. This is incorporated into the iterative model training as the ground truth. The I2I translation process is designed with two constraints: (1) a convexity consistency constraint that relies on Hessian analysis to preserve the geometric properties and (2) an intensity distribution consistency constraint based on Kullback-Leibler divergence to preserve the intensity distribution of blobs. BlobCUT learns the inherent noise distribution from the target noisy blob images and performs image translation from the noisy domain to the clean domain, effectively functioning as a denoising process to support blob identification. To validate the performance of BlobCUT, we evaluate it on a 3D simulated dataset of blobs and a 3D MRI dataset of mouse kidneys. We conduct a comparative analysis involving six state-of-the-art methods. Our findings reveal that BlobCUT exhibits superior performance and training efficiency, utilizing only 56.6% of the training time required by the state-of-the-art BlobDetGAN. This underscores the effectiveness of BlobCUT in accurately segmenting small blobs while achieving notable gains in training efficiency.

Synthesis and Expression of a Targeted, Ferritin-Based Tracer for PET Imaging of Kidney Glomeruli.

Cationic ferritin (CF) has been developed as a multimodal, targeted imaging tracer to directly detect and map nephrons in the kidney in vivo. Direct detection of functional nephrons provides a unique, sensitive biomarker to predict or monitor kidney disease progression. CF has been developed to map functional nephron number from magnetic resonance imaging (MRI) or positron emission tomography (PET). Previous preclinical imaging studies have used non-human-derived ferritin and commercial formulations that must still be developed for translation to clinical use. Here we describe the reproducible formulation of CF (either derived from horse or from human recombinant ferritin) optimized for intravenous injection and radiolabeling by PET. The human recombinant heteropolymer ferritin is spontaneously assembled in liquid culture (Escherichia coli, E. coli) and modified to form human recombinant cationic ferritin (HrCF) to mitigate potential immunologic reactions for use in humans.

Use of novel structural features to identify urinary biomarkers during acute kidney injury that predict progression to chronic kidney disease.

BACKGROUND: A significant barrier to biomarker development in the field of acute kidney injury (AKI) is the use of kidney function to identify candidates. Progress in imaging technology makes it possible to detect early structural changes prior to a decline in kidney function. Early identification of those who will advance to chronic kidney disease (CKD) would allow for the initiation of interventions to halt progression. The goal of this study was to use a structural phenotype defined by magnetic resonance imaging and histology to advance biomarker discovery during the transition from AKI to CKD. METHODS: Urine was collected and analyzed from adult C57Bl/6 male mice at four days and 12 weeks after folic acid-induced AKI. Mice were euthanized 12 weeks after AKI and structural metrics were obtained from cationic ferritin-enhanced-MRI (CFE-MRI) and histologic assessment. The fraction of proximal tubules, number of atubular glomeruli (ATG), and area of scarring were measured histologically. The correlation between the urinary biomarkers at the AKI or CKD and CFE-MRI derived features was determined, alone or in combination with the histologic features, using principal components. RESULTS: Using principal components derived from structural features, twelve urinary proteins were identified at the time of AKI that predicted structural changes 12 weeks after injury. The raw and normalized urinary concentrations of IGFBP-3 and TNFRII strongly correlated to the structural findings from histology and CFE-MRI. Urinary fractalkine concentration at the time of CKD correlated with structural findings of CKD. CONCLUSIONS: We have used structural features to identify several candidate urinary proteins that predict whole kidney pathologic features during the transition from AKI to CKD, including IGFBP-3, TNFRII, and fractalkine. In future work, these biomarkers must be corroborated in patient cohorts to determine their suitability to predict CKD after AKI.

Mapping single-nephron filtration in the isolated, perfused rat kidney using magnetic resonance imaging.

The kidney has an extraordinary ability to maintain glomerular filtration despite natural fluctuations in blood pressure and nephron loss. This is partly due to local coordination between single-nephron filtration and vascular perfusion. An improved understanding of the three-dimensional (3-D) functional coordination between nephrons and the vasculature may provide a new perspective of the heterogeneity of kidney function and could inform targeted therapies and timed interventions to slow or prevent the progression of kidney disease. Here, we developed magnetic resonance imaging (MRI) tools to visualize single-nephron function in 3-D throughout the isolated perfused rat kidney. We used an intravenous slow perfusion of a glomerulus-targeted imaging tracer [cationized ferritin (CF)] to map macromolecular dynamics and to identify glomeruli in 3-D, followed by a bolus of a freely filtered tracer (gadolinium diethylenetriamine penta-acetic acid) to map filtration kinetics. There was a wide intrakidney distribution of CF binding rates and estimated single-nephron glomerular filtration rate (eSNGFR) between nephrons. eSNGFR and CF uptake rates did not vary significantly by distance from the kidney surface. eSNGFR varied from ∼10 to ∼100 nL/min throughout the kidney. Whole single-kidney GFR was similar across all kidneys, despite differences in the distributions eSNGFR of and glomerular number, indicating a robust adaptive regulation of individual nephrons to maintain constant single-kidney GFR in the presence of a natural variation in nephron number. This work provides a framework for future studies of single-nephron function in the whole isolated perfused kidney and experiments of single-nephron function in vivo using MRI.NEW & NOTEWORTHY We report MRI tools to measure and map single-nephron function in the isolated, perfused rat kidney. We used imaging tracers to identify nephrons throughout the kidney and to measure the delivery and filtration of the tracers at the location of the glomeruli. With this technique, we directly measured physiological parameters including estimated single-nephron glomerular filtration rate throughout the kidney. This work provides a foundation for new studies to simultaneously map the function of large numbers of nephrons.

The relationship between diffusion heterogeneity and microstructural changes in high-grade gliomas using Monte Carlo simulations.

PURPOSE: Diffusion-weighted imaging (DWI) may aid accurate tumor grading. Decreased diffusivity and increased diffusion heterogeneity measures have been observed in high-grade gliomas using the non-monoexponential models for DWI. However, DWI measures concerning tissue characteristics in terms of pathophysiological and structural changes are yet to be established. Thus, this study aims to investigate the relationship between the diffusion measurements and microstructural changes in the presence of high-grade gliomas using a three-dimensional Monte Carlo simulation with systematic changes of microstructural parameters. METHODS: Water diffusion was simulated in a microenvironment along with changes associated with the presence of high-grade gliomas, including increases in cell density, nuclear volume, extracellular volume (VFex), and extracellular tortuosity (λex), and changes in membrane permeability (Pmem). DWI signals were simulated using a pulsed gradient spin-echo sequence. The sequence parameters, including the maximum gradient strength and diffusion time, were set to be comparable to those of clinical scanners and advanced human MRI systems. The DWI signals were fitted using the gamma distribution and diffusional kurtosis models with b-values up to 6000 and 2500 s/mm2, respectively. RESULTS: The diffusivity measures (apparent diffusion coefficients (ADC), Dgamma of the gamma distribution model and Dapp of the diffusional kurtosis model) decreased with increases in cell density and λex, and a decrease in Pmem. These diffusivity measures increased with increases in nuclear volume and VFex. The diffusion heterogeneity measures (σgamma of the gamma distribution model and Kapp of the diffusional kurtosis model) increased with increases in cell density or nuclear volume at the low Pmem, and a decrease in Pmem. Increased σgamma was also associated with an increase in VFex. CONCLUSION: Among simulated microstructural changes, only increases in cell density at low Pmem or decreases in Pmem corresponded to both the decreased diffusivity and increased diffusion heterogeneity measures. The results suggest that increases in cell density at low Pmem or decreases in Pmem may be associated with the diffusion changes observed in high-grade gliomas.

Estimating Nephron Number from Biopsies: Impact on Clinical Studies.

BACKGROUND: Accumulating evidence supports an association between nephron number and susceptibility to kidney disease. However, it is not yet possible to directly measure nephron number in a clinical setting. Recent clinical studies have used glomerular density from a single biopsy and whole kidney cortical volume from imaging to estimate nephron number and single nephron glomerular filtration rate. However, the accuracy of these estimates from individual subjects is unknown. Furthermore, it is not clear how sample size or biopsy location may influence these estimates. These questions are critical to study design, and to the potential translation of these tools to estimate nephron number in individual subjects. METHODS: We measured the variability in estimated nephron number derived from needle or virtual biopsies and cortical volume in human kidneys declined for transplantation. We performed multiple needle biopsies in the same kidney, and examined the three-dimensional spatial distribution of nephron density by magnetic resonance imaging. We determined the accuracy of a single-kidney biopsy to predict the mean nephron number estimated from multiple biopsies from the same kidney. RESULTS: A single needle biopsy had a 15% chance and virtual biopsy had a 60% chance of being within 20% of the whole-kidney nephron number. Single needle biopsies could be used to detect differences in nephron number between large cohorts of several hundred subjects. CONCLUSIONS: The number of subjects required to accurately detect differences in nephron number between populations can be predicted on the basis of natural intrakidney variability in glomerular density. A single biopsy is insufficient to accurately predict nephron number in individual subjects.

Premature differentiation of nephron progenitor cell and dysregulation of gene pathways critical to kidney development in a model of preterm birth.

Preterm birth is a leading cause of neonatal morbidity. Survivors have a greater risk for kidney dysfunction and hypertension. Little is known about the molecular changes that occur in the kidney of individuals born preterm. Here, we demonstrate that mice delivered two days prior to full term gestation undergo premature cessation of nephrogenesis, resulting in a lower glomerular density. Kidneys from preterm and term groups exhibited differences in gene expression profiles at 20- and 27-days post-conception, including significant differences in the expression of fat-soluble vitamin-related genes. Kidneys of the preterm mice exhibited decreased proportions of endothelial cells and a lower expression of genes promoting angiogenesis compared to the term group. Kidneys from the preterm mice also had altered nephron progenitor subpopulations, early Six2 depletion, and altered Jag1 expression in the nephrogenic zone, consistent with premature differentiation of nephron progenitor cells. In conclusion, preterm birth alone was sufficient to shorten the duration of nephrogenesis and cause premature differentiation of nephron progenitor cells. These candidate genes and pathways may provide targets to improve kidney health in preterm infants.

Image analysis techniques to map pyramids, pyramid structure, glomerular distribution, and pathology in the intact human kidney from 3-D MRI.

Kidney pathologies are often highly heterogeneous. To comprehensively understand kidney structure and pathology, it is critical to develop tools to map tissue microstructure in the context of the whole, intact organ. Magnetic resonance imaging (MRI) can provide a unique, three-dimensional view of the kidney and allows for measurements of multiple pathological features. Here, we developed a platform to systematically render and map gross and microstructural features of the human kidney based on three-dimensional MRI. These features include pyramid number and morphology as well as the associated medulla and cortex. In a subset of these kidneys, we also mapped individual glomeruli and glomerular volumes using cationic ferritin-enhanced MRI to report intrarenal heterogeneity in glomerular density and size. Finally, we rendered and measured regions of nephron loss due to pathology and individual glomerular volumes in each pyramidal unit. This work provides new tools to comprehensively evaluate the kidney across scales, with potential applications in anatomic and physiological research, transplant allograft evaluation, biomarker development, biopsy guidance, and therapeutic monitoring. These image rendering and analysis tools could eventually impact the field of transplantation medicine to improve longevity matching of donor allografts and recipients and reduce discard rates through the direct assessment of donor kidneys.NEW & NOTEWORTHY We report the application of cutting-edge image analysis approaches to characterize the pyramidal geometry, glomerular microstructure, and heterogeneity of the whole human kidney imaged using MRI. This work establishes a framework to improve the detection of microstructural pathology to potentially facilitate disease monitoring or transplant evaluation in the individual kidney.

Mapping kidney tubule diameter ex vivo by diffusion MRI.

Tubular pathologies are a common feature of kidney disease. Current metrics to assess kidney health, in vivo or in transplant, are generally based on urinary or serum biomarkers and pathological findings from kidney biopsies. Biopsies, usually taken from the kidney cortex, are invasive and prone to sampling error. Tools to directly and noninvasively measure tubular pathology could provide a new approach to assess kidney health. This study used diffusion magnetic resonance imaging (dMRI) as a noninvasive tool to measure the size of the tubular lumen in ex vivo, perfused kidneys. We first used Monte Carlo simulations to demonstrate that dMRI is sensitive to restricted tissue water diffusion at the scale of the kidney tubule. We applied dMRI and biophysical modeling to examine the distribution of tubular diameters in ex vivo, fixed kidneys from mice, rats, and a human donor. The biophysical model to fit the dMRI signal was based on a superposition of freely diffusing water and water diffusing inside infinitely long cylinders of different diameters. Tubular diameters measured by dMRI were within 10% of those measured by histology within the same tissue. Finally, we applied dMRI to investigate kidney pathology in a mouse model of folic-acid-induced acute kidney injury. dMRI detected heterogeneity in the distribution of tubules within the kidney cortex of mice with acute kidney injury compared with control mice. We conclude that dMRI can be used to measure the distribution of tubule diameters in the kidney cortex ex vivo and that dMRI may provide a new noninvasive biomarker of tubular pathology.NEW & NOTEWORTHY Tubular pathologies are a common feature of kidney disease. Current metrics to assess kidney health, in vivo or in transplant, are generally based on urinary or serum biomarkers and pathological findings from kidney biopsies. Diffusion MRI can be used to measure the distribution of tubule diameters in the kidney cortex ex vivo and may provide a new noninvasive biomarker of tubular pathology.

Dynamic Contrast Enhancement (DCE) MRI-Derived Renal Perfusion and Filtration: Basic Concepts.

Dynamic contrast-enhanced (DCE) MRI monitors the transit of contrast agents, typically gadolinium chelates, through the intrarenal regions, the renal cortex, the medulla, and the collecting system. In this way, DCE-MRI reveals the renal uptake and excretion of the contrast agent. An optimal DCE-MRI acquisition protocol involves finding a good compromise between whole-kidney coverage (i.e., 3D imaging), spatial and temporal resolution, and contrast resolution. By analyzing the enhancement of the renal tissues as a function of time, one can determine indirect measures of clinically important single-kidney parameters as the renal blood flow, glomerular filtration rate, and intrarenal blood volumes. Gadolinium-containing contrast agents may be nephrotoxic in patients suffering from severe renal dysfunction, but otherwise DCE-MRI is clearly useful for diagnosis of renal functions and for assessing treatment response and posttransplant rejection.Here we introduce the concept of renal DCE-MRI, describe the existing methods, and provide an overview of preclinical DCE-MRI applications to illustrate the utility of this technique to measure renal perfusion and glomerular filtration rate in animal models.This publication is based upon work from the COST Action PARENCHIMA, a community-driven network funded by the European Cooperation in Science and Technology (COST) program of the European Union, which aims to improve the reproducibility and standardization of renal MRI biomarkers. This introduction is complemented by two separate publications describing the experimental procedure and data analysis.

Analysis Protocol for Dynamic Contrast Enhanced (DCE) MRI of Renal Perfusion and Filtration.

Here we present an analysis protocol for dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) data of the kidneys. It covers comprehensive steps to facilitate signal to contrast agent concentration mapping via T1 mapping and the calculation of renal perfusion and filtration parametric maps using model-free approaches, model free analysis using deconvolution, the Toft's model and a Bayesian approach.This chapter is based upon work from the COST Action PARENCHIMA, a community-driven network funded by the European Cooperation in Science and Technology (COST) program of the European Union, which aims to improve the reproducibility and standardization of renal MRI biomarkers. This analysis protocol chapter is complemented by two separate chapters describing the basic concept and experimental procedure.

Small Blob Detector Using Bi-Threshold Constrained Adaptive Scales.

Recent advances in medical imaging technology bring great promises for medicine practices. Imaging biomarkers are discovered to inform disease diagnosis, prognosis, and treatment assessment. Detecting and segmenting objects from images are often the first steps in quantitative measurement of these biomarkers. The challenges of detecting objects in images, particularly small objects known as blobs, include low image resolution, image noise and overlap among the blobs. This research proposes a Bi-Threshold Constrained Adaptive Scale (BTCAS) blob detector to uncover the relationship between the U-Net threshold and the Difference of Gaussian (DoG) scale to derive a multi-threshold, multi-scale small blob detector. With lower and upper bounds on the probability thresholds from U-Net, two binarized maps of the distance are rendered between blob centers. Each blob is transformed to a DoG space with an adaptively identified local optimum scale. A Hessian convexity map is rendered using the adaptive scale, and the under-segmentation typical of the U-Net is resolved. To validate the performance of the proposed BTCAS, a 3D simulated dataset (n = 20) of blobs, a 3D MRI dataset of human kidneys and a 3D MRI dataset of mouse kidneys, are studied. BTCAS is compared against four state-of-the-art methods: HDoG, U-Net with standard thresholding, U-Net with optimal thresholding, and UH-DoG using precision, recall, F-score, Dice and IoU. We conclude that BTCAS statistically outperforms the compared detectors.

Magnetic resonance imaging accurately tracks kidney pathology and heterogeneity in the transition from acute kidney injury to chronic kidney disease.

Acute kidney injury (AKI) increases the risk for chronic kidney disease (CKD). However, there are few tools to detect microstructural changes after AKI. Here, cationic ferritin-enhanced magnetic resonance imaging (CFE-MRI) was applied to examine the heterogeneity of kidney pathology in the transition from AKI to CKD. Adult male mice received folic acid followed by cationic ferritin and were euthanized at four days (AKI), four weeks (CKD-4) or 12 weeks (CKD-12). Kidneys were examined by histologic methods and CFE-MRI. In the CKD-4 and CKD-12 groups, glomerular number was reduced and atubular cortical lesions were observed. Apparent glomerular volume was larger in the AKI, CKD-4 and CKD-12 groups compared to controls. Glomerular hypertrophy occurred with ageing. Interglomerular distance and glomerular density were combined with other MRI metrics to distinguish the AKI and CKD groups from controls. Despite significant heterogeneity, the noninvasive (MRI-based) metrics were as accurate as invasive (histological) metrics at distinguishing AKI and CKD from controls. To assess the toxicity of cationic ferritin in a CKD model, CKD-4 mice received cationic ferritin and were examined one week later. The CKD-4 groups with and without cationic ferritin were similar, except the iron content of the kidney, liver, and spleen was greater in the CKD-4 plus cationic ferritin group. Thus, our study demonstrates the accuracy and safety of CFE-MRI to detect whole kidney pathology allowing for the development of novel biomarkers of kidney disease and providing a foundation for future in vivo longitudinal studies in mouse models of AKI and CKD to track nephron fate.

Nephron number and its determinants: a 2020 update.

Studies of human nephron number have been conducted for well over a century and have uncovered a large variability in nephron number. However, the mechanisms influencing nephron endowment and loss, along with the etiology for the wide range among individuals are largely unknown. Advances in imaging technology have allowed investigators to revisit the principles of renal structure and physiology and their roles in the progression of kidney disease. Here, we will review the latest data on the influences impacting nephron number, innovations made over the last 6 years to understand and integrate renal structure and function, and new developments in the tools used to count nephrons in vivo.

Mapping nephron mass in vivo using positron emission tomography.

Nephron number varies widely in humans. A low nephron endowment at birth or a loss of functioning nephrons is strongly linked to increased susceptibility to chronic kidney disease. In this work, we developed a contrast agent, radiolabeled cationic ferritin (RadioCF), to map functioning glomeruli in vivo in the kidney using positron emission tomography (PET). PET radiotracers can be detected in trace doses (<30 nmol), making them useful for rapid clinical translation. RadioCF is formed from cationic ferritin (CF) and with a radioisotope, Cu-64, incorporated into the ferritin core. We showed that RadioCF binds specifically to kidney glomeruli after intravenous injection in mice, whereas radiolabeled noncationic ferritin (RadioNF) and free Cu-64 do not. We then showed that RadioCF-PET can distinguish kidneys in healthy wild-type (WT) mice from kidneys in mice with oligosyndactylism (Os/+), a model of congenital hypoplasia and low nephron mass. The average standardized uptake value (SUV) measured by PET 90 min after injection was 21% higher in WT mice than in Os/+ mice, consistent with the higher glomerular density in WT mice. The difference in peak SUV from SUV at 90 min correlated with glomerular density in male mice from both WT and Os/+ cohorts (R2 = 0.98). Finally, we used RadioCF-PET to map functioning glomeruli in a donated human kidney. SUV within the kidney correlated with glomerular number (R2= 0.78) measured by CF-enhanced magnetic resonance imaging in the same locations. This work suggests that RadioCF-PET appears to accurately detect nephron mass and has the potential for clinical translation.

Beyond the tubule: pathological variants of LRP2, encoding the megalin receptor, result in glomerular loss and early progressive chronic kidney disease.

Pathogenic variants in the LRP2 gene, encoding the multiligand receptor megalin, cause a rare autosomal recessive syndrome: Donnai-Barrow/Facio-Oculo-Acoustico-Renal (DB/FOAR) syndrome. Because of the rarity of the syndrome, the long-term consequences of the tubulopathy on human renal health have been difficult to ascertain, and the human clinical condition has hitherto been characterized as a benign tubular condition with asymptomatic low-molecular-weight proteinuria. We investigated renal function and morphology in a murine model of DB/FOAR syndrome and in patients with DB/FOAR. We analyzed glomerular filtration rate in mice by FITC-inulin clearance and clinically characterized six families, including nine patients with DB/FOAR and nine family members. Urine samples from patients were analyzed by Western blot analysis and biopsy materials were analyzed by histology. In the mouse model, we used histological methods to assess nephrogenesis and postnatal renal structure and contrast-enhanced magnetic resonance imaging to assess glomerular number. In megalin-deficient mice, we found a lower glomerular filtration rate and an increase in the abundance of injury markers, such as kidney injury molecule-1 and N-acetyl-ß-d-glucosaminidase. Renal injury was validated in patients, who presented with increased urinary kidney injury molecule-1, classical markers of chronic kidney disease, and glomerular proteinuria early in life. Megalin-deficient mice had normal nephrogenesis, but they had 19% fewer nephrons in early adulthood and an increased fraction of nephrons with disconnected glomerulotubular junction. In conclusion, megalin dysfunction, as present in DB/FOAR syndrome, confers an increased risk of progression into chronic kidney disease.

Mapping vascular and glomerular pathology in a rabbit model of neonatal acute kidney injury using MRI.

Acute kidney injury (AKI) in premature neonates is common due to the administration of life-saving therapies. The impact of AKI on renal morphology and susceptibility to further renal damage is poorly understood. Recent advances in radiological imaging have allowed integration of soft tissue morphology in the intact organ, facilitating a more complete understanding of changes in tissue microstructure associated with pathology. Here, we applied magnetic resonance imaging (MRI) to detect both glomerular and vascular changes in a rabbit model of neonatal AKI, induced by indomethacin and gentamicin. Using combined spin-echo MRI and cationic ferritin enhanced gradient-echo MRI (CFE-MRI), we observed (a) an increased cortical arterial diameter in the AKI cohort compared to healthy controls, and (b) focal loss of vascular density and glomerular loss in a circumferential band ~1 mm from the cortical surface. This combined use of vascular and glomerular imaging may give insight into the etiology of AKI and its impact on renal health later in life.

Improved small blob detection in 3D images using jointly constrained deep learning and Hessian analysis.

Imaging biomarkers are being rapidly developed for early diagnosis and staging of disease. The development of these biomarkers requires advances in both image acquisition and analysis. Detecting and segmenting objects from images are often the first steps in quantitative measurement of these biomarkers. The challenges of detecting objects in images, particularly small objects known as blobs, include low image resolution, image noise and overlap between the blobs. The Difference of Gaussian (DoG) detector has been used to overcome these challenges in blob detection. However, the DoG detector is susceptible to over-detection and must be refined for robust, reproducible detection in a wide range of medical images. In this research, we propose a joint constraint blob detector from U-Net, a deep learning model, and Hessian analysis, to overcome these problems and identify true blobs from noisy medical images. We evaluate this approach, UH-DoG, using a public 2D fluorescent dataset for cell nucleus detection and a 3D kidney magnetic resonance imaging dataset for glomerulus detection. We then compare this approach to methods in the literature. While comparable to the other four comparing methods on recall, the UH-DoG outperforms them on both precision and F-score.