Objective: Takotsubo syndrome (TTS) is an acute form of heart failure characterized by transient left ventricular systolic dysfunction. Given the complex cardiohepatic interactions observed in heart failure, this study aimed to evaluate the prognostic significance of hepatic T1 mapping in patients with TS. Materials and Methods: In this retrospective pilot study, cardiovascular magnetic resonance (CMR) including hepatic T1 mapping was performed in 66 consecutive patients with TTS (60 females; mean age 70.96 ± 10.11 years). The median duration of long-term follow-up was 7 months (interquartile range, 2-16 months). The primary endpoint was a composite of out-of-hospital all-cause mortality and major cardiovascular or cerebrovascular adverse events, including heart failure hospitalization, TTS recurrence, and ischemic stroke. Results: During the median follow-up period of 7 months, 12 (18%) patients experienced the primary endpoint. Kaplan-Meier analysis revealed a significantly lower event-free survival in patients with higher hepatic T1 values (log-rank, p = 0.001). In multivariable Cox regression analysis, hepatic T1 mapping emerged as an independent predictor of adverse outcomes (HR 1.010; 95% CI 1.002-1.017, p = 0.010). Conclusions: Elevated hepatic T1 mapping values were independently associated with an increased risk of adverse cardiovascular events during follow-up. Incorporating hepatic T1 mapping into the clinical evaluation of patients with TTS may improve risk stratification and support more personalized management strategies.

Takotsubo syndrome (TS) is characterized by transient left ventricular dysfunction accompanied by dynamic changes in myocardial tissue; however, differences in cardiac magnetic resonance (CMR) findings across disease phases remain incompletely characterized, particularly in large multicenter cohorts. This retrospective analysis from the multicenter EVOLUTION registry included 439 consecutive patients with TS (400 females; mean age 70.01 ± 11.59 years), stratified according to the time from symptom onset to CMR into acute (1 to 72 hours), subacute (4 to 21 days), and late (≥22 days) acquisition groups. Among these, 146 (33%) were classified as acute, 266 (60%) as subacute, and 27 (6%) as late. Biventricular systolic function was higher in patients imaged at later time points (both p = 0.001). Myocardial edema and late gadolinium enhancement (LGE) were more prevalent and extensive in patients imaged earlier and less evident in those imaged later. In multivariable analysis, T2-mapping Z-score and LGE extent were independently associated with earlier timing of CMR. T2-mapping Z-score decreased by approximately 0.22 units per day, corresponding to an average relative decline of 3% to 4% per day. In conclusion, cross-sectional CMR assessment in TS demonstrates that patients imaged at later time points exhibit more preserved systolic function and lower prevalence of myocardial edema and LGE, supporting the dynamic and reversible nature of myocardial injury in this condition; however, longitudinal studies with serial imaging are needed to confirm these findings.

Carotid atherosclerosis is a major contributor to ischemic stroke. While luminal stenosis has historically guided treatment decisions, growing evidence indicates that plaque composition, vascular inflammation and perivascular adipose tissue (PVAT) may be more closely linked to clinical outcomes and plaque vulnerability. This study aimed to characterize carotid PVAT using photon-counting computed tomography (PCCT) and to evaluate its spatial behavior and variability in a cohort of asymptomatic patients. We retrospectively analyzed PCCT angiography data from 20 asymptomatic patients. A custom-developed Python algorithm was used to segment concentric perivascular layers from 1 mm to 5 mm around the carotid artery. For each layer, we quantified attenuation values in Hounsfield Units (HU) and voxel counts. Statistical comparisons were performed across layers and between sides. Mean PVAT attenuation decreased progressively with increasing distance from the carotid wall. Significant differences were observed between inner and outer layers, particularly between the 1 mm and 3–5 mm annuli. Circle-by-circle analysis revealed substantial inter-individual variability in HU trends. Voxel count increased with annular thickness, but variability (SD and CV) also rose in outer layers. No significant differences were found between left and right carotid arteries in either attenuation or voxel distribution. Photon-counting CT enables detailed, layer-specific assessment of carotid PVAT. The observed attenuation patterns and inter-individual variability suggest that PVAT profiling may provide valuable insights into local vascular inflammation and plaque vulnerability. These findings support the potential of PCCT as a noninvasive tool for vascular risk stratification beyond luminal stenosis. Question Can photon-counting CT enable a reliable, layer-by-layer quantitative characterization of carotid perivascular adipose tissue in asymptomatic patients beyond luminal stenosis assessment? Findings Photon-counting CT demonstrated a progressive decrease in PVAT attenuation with increasing distance from the carotid wall and marked inter-individual variability across concentric layers. Clinical relevance Layer-specific PVAT profiling with photon-counting CT may provide a noninvasive imaging marker of local vascular inflammation, supporting improved carotid risk stratification beyond stenosis severity, even in asymptomatic individuals.

Objectives: To assess the diagnostic accuracy of a deep learning (DL)-based algorithm for non-invasive computation of fractional flow reserve (FFR-CT) from coronary computed tomography angiography (CCTA) and to evaluate the model's ability to automatically assign cardiovascular risk categories according to the Coronary Artery Disease-Reporting and Data System (CAD-RADS). Materials and Methods: Sixty patients with suspected coronary artery disease who underwent both CCTA and invasive coronary angiography (ICA) were retrospectively included in this multicenter study. Curved multiplanar reconstructions derived from CCTA were analyzed by the deep learning-based model to estimate FFR-CT values and to automatically assign CAD-RADS risk categories. The diagnostic performance of the software for the identification of hemodynamically significant coronary stenoses was evaluated using ICA as the reference standard. Receiver operating characteristic (ROC) curve analysis was performed to determine the area under the curve (AUC), sensitivity, and specificity on both a per-patient and per-vessel basis. Finally, agreement between CAD-RADS risk categories assigned by the DL algorithm and those determined by an expert radiologist was assessed. Results: FFR-CT demonstrated high diagnostic accuracy, with AUC of 0.935, sensitivity of 93.2%, specificity of 93.7%, and excellent agreement with reference standard (k = 0.836) on a per-patient level. Per-vessel diagnostic performance was consistently high across all major coronary arteries, with the left anterior descending artery (LAD) showing the highest accuracy (AUC = 0.932). Automated CAD-RADS classifications generated by the software showed good agreement with those assigned by human (k = 0.765). Conclusions: The DL-based model demonstrated high diagnostic accuracy and represents a promising noninvasive approach for ischemia assessment and cardiovascular risk stratification.

BACKGROUND. Late gadolinium enhancement (LGE) has traditionally been considered absent in Takotsubo syndrome (TS). However, accumulating evidence indicates the finding's presence during the acute phase in a subset of patients. OBJECTIVE. The purpose of this study was to evaluate the frequency of LGE, identify factors associated with LGE presence, assess prognostic implications of LGE, and compare methods for quantifying LGE extent, in patients with TS undergoing cardiac MRI. METHODS. This retrospective study included 370 patients (338 women and 42 men; mean age 69.7 ± 12.0 [SD] years) from the nine-center EVOLUTION (Exploring the Evolution in Prognostic Capability of Multi-Sequence Cardiac Magnetic Resonance in Patients Affected by Takotsubo Cardiomyopathy) registry from November 21, 2007, to December 22, 2024. The registry included patients with hospital admission for TS who underwent cardiac MRI within 10 days after symptom onset; patients were required to fulfill professional society criteria for TS diagnosis. Two radiologists independently reviewed LGE images to assess examinations for the visual presence of LGE, resolving discrepancies for further analyses. In patients with LGE, a radiologist quantified LGE extent visually and using semiautomated methods (2-SD, 3-SD, and 5-SD threshold methods relative to remote myocardial signal intensity; full width at half-maximum method relative to LGE peak signal intensity). In-hospital adverse events (death or major cardiac or cerebrovascular events) were identified. RESULTS. The two radiologists identified LGE in 58 (15.7%) and 54 (14.6%) patients; by consensus, LGE was present in 58 (15.7%) patients. In multivariable analysis, LGE presence was independently associated with a shorter interval from symptom onset to cardiac MRI (OR per day = 0.81; p = .003) and a greater extent of myocardial edema on T2-weighted STIR images (OR per segment = 1.44; p < .001). The mean LGE extent by visual assessment was 25.5%. Among semiautomated methods, correlation with visual assessment of LGE extent was greatest for the 2-SD threshold method (ρ = 0.93). In-hospital adverse events occurred in 88 (23.8%) patients and were not significantly associated with LGE presence (p = .44) or extent by any method (all p > .05). CONCLUSION. LGE was identified in 15.7% of patients with TS and showed significant independent associations with greater myocardial edema extent and earlier MRI timing after presentation but was not associated with in-hospital adverse events. CLINICAL IMPACT. The results may provide useful context when radiologists encounter LGE on cardiac MRI in patients with TS.

BACKGROUND:White matter hyperintensities (WMH) are common MRI markers of cerebral small vessel disease, and systemic inflammation may contribute to their development. PURPOSE:The aim is to investigate the association between circulating inflammatory biomarkers and WMH burden. DATA SOURCES:PubMed, EMBASE, and the Cochrane Library (January 2010 to January 2025) were searched. STUDY SELECTION:Studies reporting blood inflammatory biomarkers and quantitative WMH burden on brain MRI were included. Studies not published in English, nonhuman studies, retracted publications, reviews, and reports without extractable data were excluded. Of 392 records, 309 unique records were screened; 29 studies met inclusion criteria, and 11 were included in the meta-analysis (n=8846). DATA ANALYSIS:Study and patient characteristics, WMH volume, and sample size were extracted. Methodologic quality was assessed with Methodologic Index for Nonrandomized Studies and certainty of evidence with Grading of Recommendations, Assessment, Development, and Evaluations. DATA SYNTHESIS:Random-effects meta-analysis and meta-regression were performed. Heterogeneity was substantial (I 2 = 99.9%), and the pooled mean WMH volume was 2576.4 mm³ (95% CI, 2391.4-2761.4). In univariate metaregression, age (β = 356.1; 95% CI, 125.5-586.2; P = .002) and diabetes (β = 652.4; 95% CI, 370.5-934.3; P < .001) were associated with greater WMH burden; in multivariate analysis, only diabetes remained significant (β = 442.0; 95% CI, 98.7-785.3; P = .01). C-reactive protein (CRP) was not significantly associated with WMH volume. LIMITATIONS:The overall certainty of evidence was low and between-study heterogeneity was high. CONCLUSIONS:Diabetes and aging may be associated with higher WMH burden, whereas CRP shows no significant association.

Atherosclerosis is a systemic disease of the arterial tree that often assumes variable forms across vascular territories. As a result, atherosclerotic plaques in different vascular beds differ in frequency, composition, stability and propensity to rupture. Here, we argue that this territorial variability is a manifestation of partially predictable principles. Embryological origin and developmental programming create regionally distinct smooth muscle cell lineages and extracellular matrix architectures, predisposing vascular segments to characteristic modes of injury and repair. Local hemodynamic processes and vessel structures further influence this architecture, leading to distinct atherosclerotic plaque evolution. Microvascular and macrovascular calcification contribute to these processes by amplifying local signals, including inflammation and environmental cues, and by driving the chondrogenic and osteogenic transformation of vascular cells. By synthesizing evidence on divergent atherosclerotic and calcification patterns across vascular beds, the concept of territory-adapted imaging and therapy may improve prevention and risk stratification.

BACKGROUND:Prognostic relevance of cardiac magnetic resonance (CMR) in takotsubo syndrome (TTS) is not fully elucidated. We aimed to assess the prognostic value of CMR-derived left ventricular stroke volume indexed (LVSVi) in patients with TTS. METHODS:Consecutive patients with TTS underwent CMR at median 5 days (3-7) after admission. CMR analysis was centralised, patients were categorised by LVSVi (<35 vs ≥35 mL/m²). Median follow-up was 360 days. Primary endpoint was a composite of major adverse cardiovascular events and all-cause-death, secondary endpoint was all-cause-death. RESULTS:This observational study included 376 patients (mean age 70±11 years; 9% male). Average left ventricular ejection fraction (LVEF) on admission was 43%, increasing to 48% at CMR imaging. 172 (46%) patients had CMR-derived LVSVi <35 mL/m²; these were older and showed higher prevalence of hypertension and dyspnoea and a lower LVEF at admission. On CMR, low LVSVi patients demonstrated smaller LV end-diastolic volumes, lower biventricular systolic function and larger myocardial oedema. Kaplan-Meier analyses showed higher primary and secondary endpoint rates in the low-LVSVi group (both log-rank p<0.01) and lowest event rates in patients having both high LVSVi and normal LVEF (both log-rank p<0.01). Multivariable Cox-regression analysis identified LVSVi as an independent predictor of both the primary (HR 0.96; 95% CI 0.92 to 0.98) and secondary (HR 0.95; 95% CI 0.90 to 0.99) endpoint. CONCLUSIONS:Early CMR after admission in patients with TTS highlights recovering LVEF with relatively low LVSVi. Low LVSVi independently predicted mid-term outcomes, and the presence of both high LVSVi and recovered LVEF identified a low-risk subgroup.

Abstract Background and aims Conventional CT is limited in differentiating soft-plaque components due to overlapping attenuation values and is frequently affected by calcification-related blooming artifacts that obscure vessel lumen and may overestimate stenosis. Photon-counting detector CT (PCD-CT) can provide enhanced spatial resolution or spectral discrimination compared with conventional energy-integrating CT, potentially improving carotid plaque characterization and mitigating artifact-related limitations. This study systematically evaluated image quality and plaque and stenosis assessment across spectral PCD-CT reconstructions. Methods We retrospectively included patients who underwent PCD-CT for carotid stenosis assessment. Four monoenergetic (ME; 40-55-70-85 keV), pure lumen (PL), and virtual non-contrast (VNC) reconstructions were analyzed. Signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), edge sharpness, plaque area, degree of NASCET stenosis, and attenuation of plaque components (intraplaque hemorrhage, lipid-rich necrotic core, fibrotic tissue, calcification) were compared using repeated-measures ANOVA with Bonferroni correction. Results Our study included 27 patients (16 men, age: 74.5 ± 9.1 years) with 50 plaques. CNR and SNR decreased with increasing ME levels (P < 0.001), while noise increased. PL yielded CNR and sharpness comparable to ME55, whereas VNC exhibited near-zero lumen signal. Stenosis remained consistent across ME reconstructions (44-45%, P = 1.00), but significantly reduced for PL (38%, P < 0.001). Lower-energy reconstructions (ME40/ME55) improved separation of attenuation values between lumen and the various plaque components. Conclusions Low-energy PCD-CT reconstructions enhance CNR and sharpness without affecting plaque and stenosis measurements. The PL algorithm may improve calcified plaque assessment, where conventional reconstructions may overestimate stenosis. PCD-CT shows promise for advanced plaque composition evaluation but further validation against MRI and histology is warranted. Conflict of interest Juul Bierens: nothing to disclose, Jelske Kuijpers: nothing to disclose, Luca Saba: nothing to disclose, Luc J.M. Smits: nothing to disclose, Robert J. van Oostenbrugge: nothing to disclose, Thomas Flohr: nothing to disclose, M. Eline Kooi: nothing to disclose, Alida A. Postma: nothing to disclose. Figure 1 - belongs to Methods Figure 2 - belongs to Results

Introduction Medico-legal investigations frequently rely on medical documentation, including imaging reports originally performed for diagnostic purposes in emergency settings which, for a number or reasons, may overlook medico-legal findings. Consequently, the accuracy and completeness of reports represent a limiting factor in ex-post medico-legal evaluations. However, imaging can be reassessed to retrieve potentially relevant forensic information, such as soft tissue lesions, which provide reproducible insights into trauma type, means compatibility, and event dynamics. CT scans, in particular, allow for deeper investigations using 3D reconstruction software and, potentially, 3D printer for courtroom presentations. Methods To test the hypothesis that forensic information may be lost in imaging interpretation, we retrospectively evaluated 242 head CT scans and relative reports, conducted in 2024 at the Emergency Department of the University Hospital of Cagliari, ensuring patients anonymity. An additional dataset of 200 scans with clinical indications suggesting forensic relevance (e.g., assault) was also analyzed to reproduce a cohort of typical medico-legal cases. A senior radiologist reevaluated each scan, focusing on the following radiological parameters: intra/extra-axial lesions, midline shift, bone fractures, and soft tissue injuries. The latter were classified by anatomical location (left/right/midline, anterior/posterior). Data were evaluated with univariate statistical analysis for statistical significance (McNeemar test) and agreement rate between original report and reevaluation (Cohen’s K). Results A full agreement was seen on major findings potentially life-threatening such as intracranial bleeding and fractures in both datasets. A significant discrepancy in soft tissue lesions was instead detected on across all the anatomical locations with moderate to poor agreement rates between observers. Results were superposable in the dataset of forensically relevant cases. Conclusions This single-centre experience revealed significant discrepancies in soft tissue lesions between scans and their report. The routinary reevaluation of clinical indicated imaging may add significant insights in medico-legal investigations being of major help in the criminal law system.

BACKGROUND AND PURPOSE:Embolic stroke of undetermined source (ESUS) may be associated with carotid artery plaques with <50% stenosis. Plaque vulnerability is multifactorial, possibly related to intraplaque hemorrhage (IPH), lipid-rich necrotic core, perivascular adipose tissue (PVAT), and calcifications. Machine learning (ML)-based plaque classification is increasingly popular but often limited in clinical interpretability by black-box nature. We applied an explainable ML approach, using noncalcified plaque components and calcification features with the SHapley Additive exPlanations (SHAP) framework to classify plaques as culprit or nonculprit. METHODS:This was a retrospective, cross-sectional study. Patients with unilateral anterior circulation ESUS with calcified carotid plaques in neck computed tomography (CT) angiography were analyzed. Calcification-level features were derived from manual segmentations. Plaque-level features were assessed by a neuroradiologist and by semi-automated software. Plaques were classified as culprit if ipsilateral to stroke side. Eight classifiers were benchmarked, and a gradient-boosted decision tree (CatBoost) was further tuned. SHAP explained model decisions. RESULTS:Seventy patients yielded 116 calcified plaques (270 calcifications). Model based on five plaque- and calcification-level features achieved ROC-AUC (receiver operating characteristic area under the curve) 0.79 and precision-recall-AUC 0.86, outperforming classification based on plaque thickness ≥3 mm (ROC-AUC 0.59, p = 0.04) and IPH presence (ROC-AUC 0.51, p = 0.003). SHAP identified plaque thickness and PVAT volume as the most influential features with potential thresholds of >2.6 mm and ≥112 mm3, respectively.f CONCLUSIONS: ML model trained with noncalcified plaque and calcification features can classify culprit calcified carotid plaque better than conventional criteria. Using clinically interpretable features with SHAP, the model explained its decisions and suggested hypothesis-generating thresholds.

BACKGROUND:Vascular inflammation is a key aspect of plaque vulnerability. Cross-sectional studies suggest that increased carotid perivascular adipose tissue (PVAT) attenuation on CTA, which is thought to reflect vascular inflammation, is associated with stroke. OBJECTIVES:We investigated the predictive value of carotid PVAT attenuation for ischemic stroke and TIA in a longitudinal study of symptomatic patients with carotid plaque. METHODS:We included patients with recent TIA or stroke and a ≥2 ​mm carotid plaque with <70 ​% stenosis who underwent CTA and MRI and were clinically followed-up for 5 years. Mean PVAT attenuation (-190 to -30 Hounsfield Units (HU)) was quantified within a radial distance from the outer vessel wall equal to the vessel diameter on the CTA slice containing the thickest plaque. Cox proportional hazards models assessed associations with ipsilateral stroke and TIA risk. Predictive value was compared with intraplaque hemorrhage (IPH) and the European Carotid Surgery Trial (ECST) score using the C-index. RESULTS:Among 159 patients (74 ​% men; 69 (63-73) years), 11 ischemic strokes and 10 TIAs occurred over 5.1 (3.1-5.6) years. Increased PVAT attenuation was independently associated with ischemic stroke or TIA (HR: 3.21 per 10 HU increase, 95%CI:1.70-6.05) and ischemic stroke alone (HR: 5.60, 95%CI:1.93-16.31). PVAT attenuation alone predicted ischemic stroke or TIA (C-index: 0.71, 95%CI:0.70-0.73) and ischemic stroke alone (C-index: 0.78, 95%CI:0.63-0.93). Adding PVAT attenuation improved prediction beyond IPH (C-index: 0.66-0.68 to 0.81-0.84) and the ECST score (0.64-0.75 to 0.75-0.86, respectively). CONCLUSION:In symptomatic patients, PVAT attenuation is an independent marker for ischemic stroke and TIA risk.

BACKGROUND AND PURPOSE:Stroke impairs cognition and movement. Although clinical severity and infarct volume can predict functional outcomes, variability in patient responses requires advanced structural and functional connectivity methods. Disconnection markers were tested to predict functional outcomes after acute ischemic stroke using diffusion tensor imaging. METHODS:A probabilistic approach was used to quantify brain damage from white matter (WM) disconnections affecting cortical areas, using lesion masking on a tractography atlas and parcellation of gray matter into functional network nodes. Forty-three patients with acute ischemic stroke were grouped according to functional improvement (change in modified Rankin Scale score from 3-5 at discharge to 0-2 at 3-month follow-up). Significantly different structural disconnection measures between the groups were combined into a principal component and included in a logistic regression model to evaluate prediction accuracy. Fractional anisotropy (FA), radial diffusivity (RD), axial diffusivity, and mean diffusivity of the disconnected WM tracts were analyzed. RESULTS:Baseline structural disconnections in the mid-posterior and central corpus callosum predicted poor functional outcomes at 3 months, and increased somatomotor network (SMN) disconnection severity correlated with poor recovery. Age, National Institutes of Health Stroke Scale score, and structural disconnections significantly predicted functional outcomes in logistic regression models. The first principal component analysis of the dysconnectivity measures explained 88% of the total variance and improved prediction accuracy from 53.8% to 76.9%. Differences in FA and RD in the region of interest of the corpus callosum between outcome groups were statistically significant. CONCLUSIONS:Predictive outcome markers from probabilistic structural disconnection mapping in acute stroke emphasize preserving interhemispheric corpus callosum and SMN connections.

Background and Motivation In carotid ultrasound imaging, robust plaque segmentation and measurement are key requirements for dependable diagnosis and cardiovascular disease (CVD) risk evaluation. The latest techniques based on solo single stage deep learning UNet (M1:SS-UNet) leads to inconsistency in low contrast scans. We hypothesize that variants of UNet such as transformers can be more powerful paradigms. Method Design and develop novel single-stage UNet-based transformer (M6:SSwAttSkip-XmerBot-UNet) for the far wall segmentation and intima-media thickness/plaque area measurements in Japanese diabetic cohort. We use augmented data and cross-validation protocol for performance evaluation and generalization. Performance evaluation includes 16 novel metrics, namely model GFLOPS, model size, jaccard index, performance to complexity, and training time. We benchmarked our novel transformer system against six models utilizing single and double stage attention-based configurations namely M1:SS-UNet, single-stage with attention-UNet (M2:SSwAttSkip-UNet), double-stage with attention-UNet (M3:DSwAttSkip-UNet), double-stage with attention in decoder UNet (M4:DSwAttDeco-UNet), single stage with transformer in the bottle neck layer (M5:SSTransBot-UNet) and single stage with swin-transformer in every layer(M7:SSSwin-UNet). Results The results showed that M6:SSwAttSkip-XmerBot-UNet achieved a perfect normalization score of 100%, while M7:SSSwin-UNet and M5:SSTransBot-UNet followed closely by 86.07%. These models consistently outperformed M1:SS-UNet, M2:SSwAttSkip-UNet, M3:DSwAttSkip-UNet and M4:DSwAttDeco-UNet across all the 16 metrics. The final ranking of the UNet models were: SSwAttSkip-XmerBot-UNet > SSSwin-UNet > SSTransBot-UNet > DSwAttDeco-UNet > SS-UNet > SSwAttSkip-UNet > DSwAttSkip-UNet. M6:SSwAttSkip-XmerBot-UNet demonstrates comparable effectiveness with greater efficiency, making it a strong alternative. Conclusions We conclude that transformer-based models like M6:SSwAttSkip-XmerBot-UNet provide highly accurate, reliable, and automated technique that segments and measures the risk of CVD.

BACKGROUND AND PURPOSE:Intracranial atherosclerosis (ICAS) is a major cause of stroke worldwide, highly prevalent in Asian populations. While systemic inflammation plays a well-established role in the onset and progression of extracranial atherosclerosis, its relevance to ICAS remains poorly defined. This meta-analysis aimed to evaluate the association between systemic inflammatory markers and ICAS presence and progression. METHODS:The meta-analysis was performed with a comprehensive literature search that identified 292 studies, of which 12 met inclusion criteria, encompassing 19674 patients. After stratification by three biomarkers (hs-CRP, CRP, NLR), the analysis assessed the association with ICAS presence using random-effects models to estimate effect sizes. ICAS progression analysis could not be performed due to insufficient data. RESULTS:A negligible effect size magnitude was found between ICAS presence and both hs-CRP (Cliff's Delta = 0.090, 95% CI: 0.043-0.137) and NLR (Cliff's Delta = 0.151, 95% CI 0.104-0.198). Uncertainty about effect size existence or direction was found for CRP (Cohen's d = 0.145, 95% CI: -0.072-0.361). Heterogeneity was low for NLR (I² = 39.7%), but high for hs-CRP and CRP (I² = 98.5% and 71.9%, respectively). CONCLUSIONS:Increased inflammatory biomarkers do not lead to increased ICAS presence, raising concern about the actual role of systemic inflammation in ICAS pathophysiology. Despite substantial heterogeneity, the direction and magnitude of effect sizes were consistent across biomarkers. Future multicenter, prospective studies are further needed to define the role of inflammation in ICAS progression.

Purpose To develop and evaluate RectoMap, a robust open-source deep learning pipeline for fully automatic 3D segmentation of rectal cancer and mesorectum on T2-weighted MRI, with a specific focus on generalization across MRI scanner vendors in heterogeneous multi-institutional data. Methods In this retrospective multi-institutional study, 226 patients with locally advanced rectal cancer were imaged across 19 scanner models from four vendors (Siemens, Philips, GE, Hitachi) at 1.5T and 3T. Two complementary 3D architectures (nnUNet and U-MambaBot) were trained with 5-fold cross-validation under standard and extended MRI-specific data augmentation. Four ensemble strategies based on softmax averaging and the STAPLE algorithm were evaluated on an in-distribution (ID; n = 45) and an out-of-distribution (OOD; n = 63, previously unseen vendors) test set. Segmentation performance was quantified using the Dice similarity coefficient (DSC) and the 95th-percentile Hausdorff distance. Results 226 patients (mean age, 62.7 years ± 12.7 [SD], 139 men) were analyzed. In the ID setting, two-level STAPLE aggregation with extended augmentation achieved the highest rectal cancer DSC (0.767 ± 0.119), while full-model aggregation yielded the best mesorectum DSC (0.794 ± 0.131). MRI-specific augmentation consistently improved mesorectum segmentation. In the OOD setting, despite vendor shift, full-model aggregation achieved DSC 0.798 ± 0.126 for rectal cancer and 0.776 ± 0.161 for mesorectum, demonstrating stable cross-vendor performance. Conclusion RectoMap enables accurate and robust automatic segmentation of rectal cancer and mesorectum across heterogeneous MRI data. It is publicly released as an open-source tool that can be applied directly or fine-tuned on local datasets, providing a reliable baseline even when limited institutional data are available.

Cocaine use disorder (CUD) is a significant public health problem with few treatment options. Repetitive Transcranial Magnetic Stimulation (rTMS) targeting the left dorsolateral prefrontal cortex has shown promise as a therapeutic tool for neural alterations in CUD. However, its effects on white matter (WM) microstructure and their role in treatment efficacy remain uncertain. This study aimed to assess the global impact of rTMS on WM microstructure in CUD patients. In this study, we made a longitudinal correlational tractography analysis that was conducted using Quantitative Anisotropy (QA) on diffusion MRI data from CUD patients who received either active rTMS (n = 22) or sham rTMS (n = 18) treatment. Imaging data were collected before (T0) and after two weeks of treatment (T1). Correlations were derived using nonparametric Spearman partial correlation, accounting for gender, age, and age at substance initiation through multiple regression. Tracks were selected using a p-FDR threshold of 0.05. A significant QA increase was found in 9718 tracts across the whole brain in the active rTMS group compared to the sham group, with no observed reduction in QA. The affected WM tracts included cerebellar, commissural, associative, and projective fibers, mainly in the left hemisphere. The study suggests that rTMS induces widespread changes in WM microstructure, potentially improving communication between brain regions and cognitive control in CUD patients. However, the small sample size limits the findings’ generalizability, highlighting the need for larger, longitudinal studies.