Obesity affects more than one billion people and drives cardiometabolic disease risk. While traditional definitions rely on BMI, recent criteria promote direct measurement of excess adiposity, such as magnetic resonance imaging (MRI). Using MRI-derived adipose tissue (AT) data of visceral, subcutaneous, bone marrow, cardiac, renal, hepatic, pancreatic, and skeletal muscle fat with k-means clustering, we previously identified five distinct body composition subphenotypes (I–V), each displaying unique cardiovascular risk profiles. This study aimed to establish the generalizability of these subphenotypes by replication in the German National Cohort (NAKO) and validation of their association with cardiovascular disease (CVD) risk. We analyzed cross-sectional data from 29,352 individuals (44.2% female; mean age 48±12 years; BMI 26.5±4.7 kg/m2) from the NAKO baseline examination (2014–2019), who underwent comprehensive health assessments, including interviews, questionnaires, biosample collection, and whole-body MRI. Body composition subphenotypes were replicated using a cluster validation framework. Associations with 10-year CVD risk, estimated by the Framingham score, were evaluated using linear regression. The five subphenotypes (I–V) were successfully replicated. Cluster I (“lean”) was youngest, had the lowest prevalence of hypertension, hypercholesterolemia, and diabetes, and the lowest CVD risk. This cluster was the reference category in further analyses. Cluster II (“average adiposity”) showed intermediate risk factor levels and a 2-fold higher CVD risk (95% CI 1.9–2.0). Cluster III (“bone and muscle adiposity”) included older participants (56±9 years) and showed a 3.7-fold higher risk (3.6–3.8), consistent with regular age-related changes. Cluster IV (“hepato-abdominal adiposity”) had a similar age (50±10 years) as cluster II (48±10 years) but adverse cardiometabolic features, elevated liver enzymes, and 3.4-fold higher risk (3.3–3.5). Cluster V (“general and pancreatic adiposity”) had the highest burden of comorbidities, and a 5-fold higher CVD risk (4.8–5.2). With an age (59±8 years) comparable to cluster III, it represents an unhealthy ageing pattern. In conclusion, MRI robustly identifies distinct body composition subphenotypes that capture the interplay of AT depots, potentially reflect aging pathways, and show differential CVD risk. Our results highlight the potential of AT distribution for personalized risk assessment and ageing trajectories.

Background and purpose:CT perfusion (CTP) is widely used to assess infarct core in acute stroke, yet real-world data on its reproducibility and temporal dynamics are limited. Materials and methods:We retrospectively identified patients with repeated CTP scans. Core and hypoperfusion volumes were quantified using standard thresholds (CBF <30 %, Tmax >6 s). Clinical and imaging data were reviewed to identify cases with disruptive events. We analyzed scan-to-scan differences in core volume, hypoperfusion volume, ASPECTS, and intensity metrics, including median Tmax (in hypoperfusion), CBF, and NCCT HU (in core), using Bland-Altman analysis and assessed their association with time between scans. Results:Among 32 patients with repeated CTP (26 with repeated NCCT), three were excluded due to disruptive events. In the remaining 29 cases, mean scan-to-scan differences for infarct core volume (4.8 ± 19.6 mL), hypoperfusion volume (3.86 ± 39.1 mL), and ASPECTS (-0.4 ± 1.6) indicated minimal systematic bias at the group level but substantial variability. Correlation coefficients were high (r = 0.90, 0.93, and 0.70, respectively; all p < 0.0001), and no statistically significant paired differences or association with scan interval were observed. Intensity-based metrics likewise showed minimal bias with lower variability (Tmax -0.3 ± 1.07 s; CBF -3.11 ± 7.3 %; NCCT HU -3.0 ± 4.9 %), high correlations (Tmax r = 0.87, CBF r = 0.91, NCCT HU r = 0.69; all p < 0.02), and no association with time between scans. Conclusions:Repeated CTP showed no systematic group-level scan-to-scan bias suggestive of infarct growth, while a substantial degree of variability was observed, with intensity-based metrics demonstrating lower variability than volume estimates. These findings support temporal consistency of perfusion-derived metrics at the group level and question the applicability of linear infarct growth rate (IGR) concepts to perfusion imaging, which primarily reflects a hemodynamic state rather than time-dependent tissue progression.

Abstract Background CSF loss in spontaneous intracranial hypotension (SIH) has been related to alterations in glymphatic flow, which is poorly understood in this disease. Advanced multi-shell diffusion-weighted MRI (dMRI) enables quantification of the interstitial free water fraction, serving as a possible surrogate marker for glymphatic system function in patients with SIH. Methods SIH Patients underwent dMRI before and after closure of a spinal CSF leak. The microstructural free water compartment (V-CSF) of the whole brain gray and white matter, corona radiata, amygdala, hippocampi and parahippocampal gyri was compared to 23 age-matched normal controls. Pre- and post-therapeutic volumetry encompassed the total ventricular, total gray and white matter compartments and mesial temporal structures. Results 23 SIH patients (50.3 ± 13.1 years, 15 women) were included. After leak closure, V-CSF increased in the global gray matter (mean pre 0.140 vs. mean post 0.151; p = 0.029), posterior corona radiata (mean pre 0.103 vs. mean post 0.108; p = 0.0055), hippocampi (mean pre 0.100 vs. mean post 0.105; p = 0.001), and parahippocampal gyri (mean pre 0.156 vs. mean post 0.177; p = 0.009). Compared to normal controls, V-CSF was decreased before leak closure in the hippocampi (mean pre 0.100 vs. mean NC 0.211; p = 0.0019) and posterior corona radiata (mean pre 0.103 vs. mean NC 0.118; p = 0.011). No significant change of total gray or white matter volume occurred after leak closure. Conclusion Closure of the spinal CSF leak leads to an increase of interstitial fluid in gray matter, corona radiata, hippocampi, and parahippocampal gyri, respectively. Our results suggest, that SIH patients may have less interstitial fluid in the hippocampi and posterior corona radiata compared to normal controls. Whether shifts in brain interstitial fluid in eloquent cerebral regions contribute to cognitive decline in patients with CSF loss should be topic of further research.

RATIONALE AND OBJECTIVES:Contrast-enhanced (CE) MRI provides clear corticomedullary contrast for renal compartment delineation but may be contraindicated or undesirable in routine practice. We aimed to enable automated extraction of renal imaging biomarkers from routine non-contrast-enhanced (NCE) T1-weighted MRI by transferring CE-derived compartment labels. MATERIALS AND METHODS:This retrospective single-center study (January 2017 to December 2021) included 200 patients with paired arterial-phase CE and NCE T1-weighted MRI. Cortex, medulla, and sinus were manually segmented on CE MRI and rigidly transferred to NCE MRI to provide voxel-level reference labels. A hierarchical 3D Deep Neural Patchworks (DNP) model was trained on 100 examinations (90 training/10 validation) and evaluated on an independent test set of 100 examinations using the transferred CE masks on NCE as reference. Performance was assessed using Dice similarity of segmentations and biomarker agreement using volumes (Pearson, MAE, Lin's CCC, and Bland-Altman). RESULTS:Whole-kidney segmentation Dice was 0.950 (left) and 0.953 (right). Total kidney volume showed high agreement with minimal bias (MAE 8.76 mL, 2.5% of mean; CCC 0.983; bias -1.56mL; 95% limits of agreement -28.81 to 25.69 mL). Cortex volume was modestly overestimated and medulla volume underestimated, shifting predicted compartment fractions toward cortex (74.7% vs. 72,1% in ground truth; medulla 21.5% vs. 24.3%; sinus 3.8% vs. 3.6%. Sinus volume maintained high concordance despite higher Dice dispersion. CONCLUSION:CE-supervised knowledge transfer enables accurate, well-calibrated total kidney volumetry from routine NCE MRI and supports contrast-free renal biomarker extraction in kidneys without major structural distortion.

BACKGROUND AND PURPOSE:The fate of ischemic tissue following a thromboembolic occlusion depends on the degree of hypoperfusion and the time from stroke onset to reperfusion. Little is known about the temporal dynamics of hypoperfusion and tissue damage. In this cross-sectional study, we compared how onset-to-imaging (OTI) time is associated with volumetric versus intensity-based imaging markers. MATERIALS AND METHODS:We retrospectively analyzed acute stroke imaging from 288 CT and 275 MR examinations. Hypoperfusion and infarct core were estimated using the VEOcore software based on standard thresholds (time to maximum [Tmax] > 6s, CBF <30%, or ADC <620 × 10-6 mm2/s). Tissue damage was quantified on NCCT using software-assisted Alberta Stroke Program Early CT Score (ASPECTS), and contralaterally normalized signal intensities (NCCT HU, ADC, DWI-b0, DWI-b1000, CBF, Tmax) within the estimated infarct core. Associations with OTI time were evaluated using multivariable linear regression, adjusting for age, sex, occlusion site, and MR field strength. RESULTS:Patients who underwent CT were older (77 [65-83] versus 72 [63-80] years, P < .001) and imaged earlier (86 [64-149] versus 102 [97-193] min, P < .001) than patients who underwent MRI. Hypoperfusion and infarct core volumes were larger on CT (143 [96-189] versus 90 [40-156] mL; 24 [12-48] versus 17 [8-35] mL; both P < .001). After adjustment, volumetric measures showed limited time-dependence: ASPECTS decreased by -0.33 points/h (P < .001) and ADC-core volume increased by +1.8 mL/h (P = .01), while perfusion volumes (CBF <30%, Tmax > 6s, Tmax > 10s) showed no significant change. Intensity measures changed markedly with time: NCCT intensity decreased by -1.1%/h (P < .001), ADC intensity by -0.69%/h (P = .001), whereas DWI-b0 increased by +2.3%/h and DWI-b1000 by +4.9%/h (both P <.001). Perfusion-based intensities were not significantly associated with time in either technique. CONCLUSIONS:In this cross-sectional analysis of multimodal data in acute ischemic stroke, tissue signal intensities showed stronger time-dependence than volumetric measures, supporting the view that infarct evolution reflects progressive tissue injury rather than consistent volumetric expansion. This suggests that "infarct growth rate" concepts may be more applicable to NCCT and DWI as parameters of tissue demise rather than to perfusion-based metrics.

Summary Background Anthropometric measures do not adequately capture heterogeneity in body fat distribution and corresponding cardiometabolic risk, whereas magnetic resonance imaging (MRI) enables precise differentiation and quantification of adipose tissue compartments and ectopic fat. We aimed to validate previously derived MRI-based body composition subphenotypes and their cardiometabolic risk profiles in two independent European cohorts. Methods Using deep learning–based image analysis, we quantified bone marrow, visceral, subcutaneous, cardiac, renal sinus, hepatic, skeletal muscle, and pancreatic fat in the imaging substudies of two population-based cohorts: the German National Cohort (NAKO, N=29,314, age range 19-74 years) and the UK Biobank (N=36,109, age range 40-69 years). Body composition subphenotypes, previously identified by k-means clustering, were evaluated using a rigorous statistical cluster validation framework with method-based and results-based approaches. In NAKO, cross-sectional associations between subphenotypes and estimated cardiovascular disease risk scores were examined using linear regression. In UK Biobank, longitudinal associations between subphenotypes and incident cardiometabolic outcomes, ascertained through hospital record linkage, were analysed using Cox regression. Findings All five body composition subphenotypes were robustly validated across both cohorts, and showed distinct fat distribution patterns and cardiometabolic risk profiles: I “lean”, II “average adiposity”, III “bone and muscle adiposity”, IV “hepato-abdominal adiposity”, and V “general and pancreatic adiposity”. Subphenotypes I–III showed progressive adipose tissue remodelling patterns likely reflecting ageing trajectories. The “hepato-abdominal adiposity” subphenotype showed highest risk of incident diabetes, whereas the “general and pancreatic adiposity” subphenotype showed highest overall cardiovascular disease burden and metabolic impairment. Interpretation MRI-derived body composition subphenotypes represent distinct fat distribution patterns that reflect ageing- and disease-related processes, which supports the potential of body composition phenotyping for improved cardiometabolic risk stratification and targeted prevention.

Background:The Alberta Stroke Program Early Computed Tomography Score (ASPECTS) is widely established to assess early ischemic changes on non-contrast computed tomography (NCCT) and guide treatment decisions in acute stroke. While automated ASPECTS tools are increasingly available, independent validation and the potential influence on human ratings remain important. The aim of this study was to evaluate agreement between an automated ASPECTS scoring system and expert readers and to examine whether software assistance was associated with a systematic shift in human ASPECTS scoring. Methods:We implemented an automated ASPECTS scoring system, incorporating image normalization, anatomical registration, and regional intensity analysis through hemispheric comparisons of net water uptake (NWU) values. A total of 224 cases were retrospectively analyzed. We included two expert unassisted readers (U1-2; reference reader group) and four software-assisted readers (A1-4). Agreement was assessed using intraclass correlation coefficient (ICC), mean difference (MD), mean absolute difference (MAD), and the distribution of absolute score differences (Δ=0/1/2/≥3). Unassisted vs assisted scores were compared using a Wilcoxon rank-sum test. Analyses were performed using two bootstrapped stratifications (approximately uniform and clinically representative). Results:Inter-rater agreement within the unassisted and assisted reader groups was high (uniform: unassisted ICC 0.96, MAD 0.65; assisted ICC 0.95, MAD 0.66; clinical: unassisted ICC 0.91, MAD 0.83; assisted ICC 0.88, MAD 0.89). Agreement between unassisted and assisted readings was similarly high (uniform: ICC 0.96, MAD 0.60, MD -0.02; clinical: ICC 0.90, MAD 0.80, MD -0.05), with no significant differences between unassisted and assisted scores (Wilcoxon: uniform P=0.82, clinical P=0.61). The automated output showed good agreement with human ratings, though consistently lower than inter-reader agreement (uniform: ICC 0.89, MAD 0.97-1.01, MD -0.20 to -0.22; clinical: ICC 0.76-0.77, MAD 1.18-1.27, MD -0.16 to -0.21). Sensitivity analyses supported an NWU threshold of approximately 7%. Conclusions:The automated system demonstrated good-to-moderate agreement with expert ratings and was not associated with a systematic group-level shift in ASPECTS scoring. It may support more standardized ASPECTS evaluation, particularly in settings with limited access to expert readers, while maintaining the autonomy of clinical judgment.

Background CT perfusion (CTP) imaging is a valuable tool for assessing cerebral blood flow. However, it is associated with high radiation doses, often exceeding those of non-contrast cranial CT. We evaluated the impact of different dose reduction approaches in CTP. Methods We retrospectively included patients who underwent CTP in a stroke referral center. Imaging was performed using four different scanners. The protocols were iteratively optimized by decreasing the tube current time product through lowering the tube current while maintaining a constant tube rotation time. Additionally, the temporal sampling interval was increased from 1.5 to 3.0 s. Volume CT dose index (CTDIvol) was compared across the optimization steps. Effective doses were calculated and the theoretical lifetime cancer risk was estimated. Results Radiation doses from 3812 cases with CTP (47.5% female, mean age 73 ± 14 years) were evaluated. Reducing the tube current by one third resulted in a dose decrease of 33.1%. Doubling the temporal sampling interval significantly reduced the dose by approximately half (47.3%-51.7%) across all scanners (all p < 0.001). Combining both adjustments on the same scanner led to an overall dose reduction of 67.7%. The lowest mean CTDIvol and effective dose across all scanners were 69.2 ± 0.2 mGy and 0.61 ± 0.08 mSv, respectively. The estimated additional lifetime cancer risk was reduced by up to 67.7%, with greater absolute benefits in younger patients. Conclusion Dose reduction measures allow for scanning CTP with exposure levels close to those of standard non-contrast CT.

Abstract Neurodegenerative diseases impair both gray matter and long-range white-matter pathways. Existing diffusion MRI approaches are either local or depend on predefined parcellations, limiting their ability to capture distributed network disruption. We introduce a streamline-wise framework to map axonal degeneration and derive disease-specific fiber “degeneromes” across Alzheimer’s disease (AD), Parkinson’s disease (PD), multiple system atrophy (MSA), and progressive supranuclear palsy (PSP). We analyzed diffusion microstructure imaging and T1-weighted MRI (3 T Siemens Prisma, 2018–2024) in AD (n = 81), PD (n = 177), MSA (n = 50), PSP (n = 35), and healthy controls (n = 26). The intraaxonal volume fraction, estimated via a Bayesian three-compartment model, was mapped along ~20,000 normative streamlines in MNI space using age- and sex-adjusted regression with FDR correction. Streamline-wise mapping revealed disease-specific degeneromes consistent with established pathoanatomical models: AD involved limbic and temporo-occipital pathways, PD showed commissural and posterior association involvement, MSA affected pontocerebellar and corticospinal tracts, and PSP involved the dentato-rubro-thalamic tract and superior cerebellar peduncle. These signatures supported group-level differentiation, and streamline-wise z-scores enabled intuitive single-patient visualization. Fiber degeneromes offer a connectome-informed biomarker with strong biological plausibility, discriminatory potential across neurodegenerative entities, and a clear route toward clinical single-patient reporting.

Automated segmentation of intracranial vessels in 3D black-blood T1W-MRI is feasible. This study evaluates its utility for detecting recurrent aneurysms and compares automatically labeled MRI (AL-BB-MRI) with time-of-flight MRA (TOF-MRA) and digital subtraction angiography (DSA). For model development, the basal intracranial arteries were manually labeled in 37 3D black-blood T1W-MRI examinations from 31 patients with previously endovascularly treated aneurysms. An independent, consecutive series of 84 patients with 90 aneurysms was assessed by three readers to identify recurrences in AL-BB-MRI and TOF-MRA. DSA served as the reference standard when available. The model achieved adequate vessel segmentation (Dice = 0.81). Agreement between AL-BB-MRI and TOF-MRA was in 70/90 (78

CT-Perfusion (CTP) is an essential part of stroke imaging. Incomplete coverage of the contrast bolus in CTP can lead to errors in post-processing that might hamper the identification of the infarct core or tissue at risk. However, the arrival of the contrast bolus depends on various technical and patient individual factors. This study investigated whether timing information from CT-angiography (CTA) can be used to optimize bolus coverage in CTP. We retrospectively reviewed cases with a multimodal stroke protocol for suspected ischemic stroke. Information on the contrast injection timing of CTA and CTP was extracted from the DICOM headers. Bolus arrival information were obtained from the CTP scan including peak time, height, and width and correlated with patient age and ejection fraction (the latter available in n = 868). The contrast timing information of the CTA was used to simulate optimized CTP timing. A total of 1,843 cases were included. CTP bolus peak position was associated with peak width (Pearsons’s r = 0.89, p < 0.001), age (Pearsons’s r = 0.40, p < 0.001), ejection fraction (Pearsons’s r=-0.25, p < 0.001), and time to scan initiation based on triggering in CTA (Pearsons’s r = 0.83, p < 0.001). Using information of the CTA timing to adjust the CTP timing, the variance of the AIF peak could significantly be reduced (p < 0.001). Our data indicate that patient individual characteristics lead to substantial variances in the contrast bolus arrival which could hamper CTP analysis. To ensure optimized coverage of the contrast bolus. CTP timing can significantly and safely be improved using timing information of preceding CTA.

BACKGROUND AND PURPOSE:White matter hyperintensities (WMH) of presumed vascular origin describe structural alterations of cerebral white matter, thought to result from cerebral small vessel disease. However, the in vivo effects of WMH on normal-appearing white matter microvasculature remain elusive. Therefore, we conducted an exploratory investigation of microvascular density in normal-appearing and pathological white matter in patients with WMH. METHODS:Using magnetic resonance imaging-based vessel size imaging we investigated an index of microvessel density in vivo in two clinical cohorts with ischaemic events (cohort_A, N = 88, mean age = 77.18) and intracranial neoplasms (cohort_B, N = 58, mean age = 65.45). For analysis, regions of interest were created for the whole normal-appearing white matter, the normal-appearing centrum semiovale, the WMH, the WMH penumbra, and the normal-appearing striatum. The severity of WMH-burden was quantified using the Wahlund score. RESULTS:In both cohorts, the index of microvessel density in the striatum was significantly higher than in normal-appearing white matter. There was no significant difference between WMH and WMH penumbra. However, both WMH and WMH penumbra had a significantly lower index of microvessel density than normal-appearing white matter in the subgroup of patients with high Wahlund scores. Lastly, the index of microvessel density in the normal-appearing centrum semiovale was higher in patients with high compared to low Wahlund scores in both cohorts. This comparison was not significant after adjusting for age. CONCLUSIONS:Our results suggest a complex relationship between cerebral small vessel disease-related WMH and microvascular changes in the normal-appearing white matter, potentially indicative of WMH-related angiogenesis.

BACKGROUND:Hypertension is closely associated with autonomic dysfunction. The role of the structural integrity of the central autonomic network (CAN) of the brain is insufficiently explored. Large-scale imaging data were used to investigate the relationship between the microstructural properties of the CAN with blood pressure (BP) and hypertension. METHODS:We analysed MRI data from 43 994 individuals to investigate whether BP levels were linked to the microstructural integrity of white matter tracts involved in autonomic control (CAN tracts). To test the specificity of these associations, we compared them to randomly selected white matter regions not specifically tied to the autonomic network, aiming to identify whether CAN tracts had a stronger connection to BP and which subsystems were particularly affected. RESULTS:Our findings showed that BP was more strongly linked to the microstructural integrity of CAN tracts than to other white matter regions. Further analysis revealed that specific CAN subsystems had distinct associations with BP, with higher levels of free water in these regions being associated with increased BP and hypertension. Additionally, the severity of hypertension was associated with the level of microstructural integrity in CAN tracts. CONCLUSION:This study provides evidence of a specific relationship between BP levels and the microstructural integrity of the CAN. We found that, particularly in cortical parts of the CAN, higher levels of free water - indicating tissue not actively involved in neural signalling - were associated with elevated BP levels and a greater risk of hypertension. This evidence supports a close link between the central autonomic system and BP from a population-imaging perspective.

Acute stroke management is time-sensitive, making time data crucial for both research and quality management. However, these time data are often not reliably captured in routine clinical practice. In this proof-of-concept study we analysed image-based time data automatically captured in the DICOM format. We enrolled data from two separate stroke centers (n = 3136 and n = 2089). Data from the first center was additionally separated into groups with large-vessel-occlusion (LVO, n = 1.092), medium-vessel-occlusions (MVO, n = 416), and no occlusion (NVO, n = 1630). The DICOM-tag StudyTime was used to analyze the distribution of scan times throughout the day. Additionally, manually documented onset- and admission were extracted from the patients’ records in a subset of cases (n = 347). Timestamps were compared across centers and occlusion groups, and a probabilistic model was developed to illustrate and compare stroke occurrence patterns throughout the day. The temporal distribution of the scan times at both centers was exceptionally consistent with a peak around noon and a nighttime low. The groups with vessel occlusions showed an earlier peak compared to those without (p < 0.04). The median interval between admission and scan time was 23 min, while the median onset-to-imaging time was 1 h:54 min. This proof-of-concept study indicates that DICOM-timestamps can reveal insights into the temporal patterns of stroke imaging and may be a promising tool for quality control and stroke research in general since they are always automatically captured by imaging devices as opposed to manual data collection in routine clinical practice.

Background and purposeA proportion of individuals recovering from COVID-19 continue to experience persistent symptoms, including fatigue and cognitive difficulties - a syndrome commonly referred to as Post-COVID condition (PCC), which affects an estimated 2-10% of cases. In this study, we evaluated cerebral blood flow (CBF) to better understand the pathophysiological mechanisms underlying PCC.Materials and methodsIn this prospective, monocentric study, we analyzed clinical and cerebral blood flow (CBF) data from a cohort of 55 patients who met the WHO diagnostic criteria for Post-COVID condition (PCC) and underwent MRI approximately 11 months after a positive PCR test for SARS-CoV-2. These PCC patients were compared to a matched control group of 36 individuals who had contracted COVID-19 but did not develop PCC. CBF was assessed using arterial spin labeling (ASL), a promising non-invasive technique that provides high spatial resolution for quantifying cerebral blood flow. Additionally, we examined changes in gray matter volume and atrophy using FreeSurfer-based cortical morphometry. We further explored the relationship between regional CBF alterations and clinical symptoms, including cognitive and olfactory function, as well as fatigue.ResultsIn our cohort, 59% of PCC patients could not return to their previous level of independence or employment due to symptoms, and 81% reported fatigue on the WEIMuS questionnaire. Conventional MRI showed no evidence of cortical atrophy. While no significant differences in regional CBF emerged after FDR correction, a more explorative threshold (p < 0.005) revealed reduced CBF in the right angular and middle occipital gyri in PCC patients. Fatigue, as assessed by the WEIMuS, was significantly correlated with reduced CBF in the right occipital regions, particularly for physical fatigue, but no associations were found with cognitive or olfactory performance.ConclusionIn PCC patients, fatigue was associated with reduced perfusion in right-sided occipital regions, suggesting a potential pathophysiological basis for this symptom. These findings may also provide an imaging biomarker to aid in the diagnosis of PCC.

Introduction Chronic kidney disease (CKD) is defined as sustained abnormalities in kidney function or structure. Genetic studies of CKD have largely focused on kidney function markers such as estimated glomerular filtration rate (eGFR). We hypothesized that genome-wide association studies (GWAS) of magnetic resonance imaging (MRI)-based kidney sub-volumes could provide insights into CKD risk genes complementary to the study of eGFR. Methods Total kidney volume (TKV) and sub-volumes for cortex, medulla, and sinus were derived from abdominal MRIs of 38,816 United Kingdom Biobank participants of European ancestry using a trained convolutional neural network. GWAS was performed for body surface area-normalized kidney volumes and eGFR for comparison. Potentially causal genes at each locus were prioritized using a developed annotation pipeline. We assessed locus overlap between volumes, biomarker-based kidney function, and clinical traits using colocalization analyses. Annotated genes were further characterized through enrichment analyses, molecular and clinical annotations, including a screen for rare, putative loss-of-function variants. Results GWAS for 9,803,932 common genetic variants identified 34 significant loci for TKV, 24 for medulla, 26 for cortex, and 71 for sinus, compared to 32 for eGFR. Prioritized genes for cortex and medulla volumes showed corresponding tissue-specific expression and were enriched for kidney development- and hypoxia-related pathways. Genetic effect sizes of significant index single nucleotide polymorphisms for TKV, cortex, and medulla volumes correlated positively with those for eGFR. Some loci such as PKHD1 and BICC1 were strongly associated with kidney volumes but not eGFR. Integration with disease information revealed that rare, putative loss-of-function variants in BICC1, and common variants with regulatory potential, are associated with increased risk for CKD and dialysis, which was not identified in a previous GWAS of eGFR Conclusions Our investigation shows that genetic findings of kidney structure can complement kidney function studies and reveal previously unrecognized CKD risk genes in the population.

Genome-wide association studies (GWAS) of kidney function typically rely on biochemical biomarkers like serum creatinine or cystatin C to estimate the glomerular filtration rate (eGFR). These biomarkers have limitations, including varying eGFR equations, biological variation, and difficulty distinguishing loci affecting kidney function from those influencing biomarker metabolism. This study explores the genetic architecture of imaging-derived kidney volumes as alternative markers of kidney function. We utilized the two-point Dixon sequence from abdominal MRIs of 38,816 UK Biobank (UKB) participants of European ancestry. Kidney compartment volumes (total kidney volume (TKV), cortex, medulla, and sinus volumes) were derived from automatically segmented kidney images using Patchwork, which uses a convolutional neural network to train ground truth data established by experienced radiologists. The volumes were normalized to body surface area and an inverse normal transformation of ranks was applied. GWAS was based on imputed UKB genotypes (allele frequency > 1%) and was performed using linear regression as implemented in the regenie software (v3.2.9). Analyses were adjusted for age, age-squared, sex, assessment centre, and the first 10 genetic principal components. We conducted a parallel eGFR GWAS for the same population and calculated genetic effect size correlations between the index SNPs for kidney volumes and eGFR. Causal genes for each locus were prioritized using a developed annotation pipeline incorporating gene proximity, Ensembl Variant Effect Predictor, genetic colocalization with cis-expression and -protein quantitative trait loci (QTL), and linkage disequilibrium. We assessed locus overlap between volumes, marker-based kidney function, and clinical traits using colocalization analyses (genepicoloc package), performed gene enrichment analysis to identify tissue-specific (GTEx v8) and renal cell type-specific (KPMP) gene expression patterns, and Gene Ontology enriched terms. Associations with renal outcomes were evaluated using results from the AstraZeneca PheWAS Portal, a platform that provides phenome-wide association study data derived from UKB exome sequencing. GWAS identified 34 significant loci for TKV, 24 for medulla, 26 for cortex, and 71 for sinus volumes (P-value < 5e-8). Pairwise colocalization analysis identified nine loci specific to the medulla and 66 to the sinus (probability of colocalization > 0.8). We observed strong correlations between genetic effect sizes of eGFR GWAS and kidney volumes for TKV (r = 0.89), cortex (r = 0.86), and medulla (r = 0.82), but a lower correlation with sinus (r = 0.37). Enrichment analyses highlighted that genes associated with cortex volume were highly expressed in the kidney cortex, while sinus volume genes were predominantly expressed in adipose and vascular tissues, supported by cell type-specific overrepresentation analyses. Gene Ontology enrichment showed that TKV and cortex genes are involved in renal development, while medulla genes are involved in hypoxia-related processes. Additionally, colocalization analysis identified 94 shared genetic signals between kidney volumes and traditional kidney function traits, supporting a shared genetic basis. The eGFR-decreasing G allele at rs77924615 (UMOD) was also associated with lower TKV (P-value = 1.5e-21). At the same time, many kidney volume-associated loci established in this study were not previously linked to traditional markers even in large GWAS meta-analyses. Moreover, we found evidence for a shared genetic basis between kidney volumes and hypertension, cardiovascular diseases, diabetes. By mining the AstraZeneca PheWAS Portal we found links between the rare, loss-of-function variants in prioritized genes (e.g., UMOD, SLC22A7) and renal outcomes: chronic kidney disease stages and kidney function markers. This study reveals novel genetic determinants of kidney structure unreported in previous biomarker-based studies, providing insights into the genetic architecture of kidney structure and its relationship with renal function.