European Radiology最新文献

Objectives: This study aimed to evaluate the diagnostic accuracy of MRI-based Node Reporting and Data System (Node-RADS) in diagnosing lymph node metastasis (LNM) and to investigate its prognostic significance in rectal cancer (RC) patients.

Materials and methods: Patients with RC who underwent radical rectal resection (including LN dissection) without any prior anti-tumour therapy between May 2019 and April 2023 were retrospectively included. Two radiologists independently scored lymph nodes using the MRI-based Node-RADS. The diagnostic performance of Node-RADS was estimated using the area under receiver operating characteristic (ROC) curves (AUC) and compared with size criteria and MRI reports conducted by experienced radiologists. Intra- and inter-observer agreement were both assessed. Disease-free survival (DFS), which served as a key postoperative prognostic indicator, was evaluated and compared between patients with low (1-3) and high (4-5) scores.

Results: Overall, 163 patients with RC were enrolled, including 53 with LNM. There were 98 men and 65 women with a mean age of 62.6 ± 10.1 years. Node-RADS showed a larger AUC (0.912) with higher sensitivity (81%) and specificity (97.3%) compared to size criteria (75.5% and 71.8%) and MRI reports (91.0% and 41.8%). Node-RADS scores were also evaluated and correlated with the prognosis in patients who had undergone radical rectal resection. A multivariable Cox model combining Node-RADS and extramural venous invasion (EMVI) showed good predictive performance for DFS (C-index: 0.718).

Conclusion: The Node-RADS scoring system, based on MRI, enhanced both sensitivity and specificity in detecting LNM in RC patients who directly received radical rectal resection and showed potential prognostic significance for RCs.

Key points: Question What is the clinical utility of the MRI-based Node-RADS for treatment-naive RC? Findings The MRI-based Node-RADS demonstrated better diagnostic accuracy (AUC 0.912) for LNM and provided significant prognostic value for DFS in treatment-naive RC patients. Clinical relevance The MRI-based Node-RADS is a reliable diagnostic method for lymph node assessment in treatment-naive RC with higher sensitivity and specificity. It is also useful in predicting postoperative outcomes.

Objectives: Preclinical declines in body mass index (BMI) are linked to accelerated Alzheimer's disease (AD) neurodegeneration and mortality, yet the temporal relationship between premorbid BMI trajectories and AD neuropathology remains unclear. This study aims to characterize stage-specific BMI dynamics preceding mild cognitive impairment (MCI)/AD diagnosis and evaluate their bidirectional associations with core AD pathologies.

Materials and methods: This longitudinal cohort study analyzed 1570 participants (mean age 73.2 ± 6.9 years; 53% male) from the Alzheimer's Disease Neuroimaging Initiative, applied linear mixed-effect models to construct BMI trajectories, and used partial correlation analysis and cross-lagged panel model to assess bidirectional associations between BMI changes and pathological progression, including β-amyloid (Aβ), tau, and neurodegeneration.

Results: Frontotemporal Aβ deposition preceded and predicted preclinical BMI decline (β = -5.74, p = 0.003), which subsequently correlated with accelerated neurodegeneration during MCI transition, including hypometabolism (r = 0.42, p < 0.001) and gray matter atrophy (r = 0.24, p = 0.01). Post-MCI diagnosis, BMI trajectories stabilized, yet lower BMI was associated with elevated cerebrospinal fluid tau levels, regardless of AD conversion. Importantly, lower premorbid BMI at MCI diagnosis was linked to faster temporo-occipital tau accumulation (r = -0.53, p = 0.01) and temporal hypometabolism (r = 0.23, p = 0.002) during MCI-to-AD progression.

Conclusions: This study suggests a temporal relationship between BMI trajectories and AD pathology: early Aβ deposition predicts preclinical BMI decline, which exacerbates tauopathy and neurodegeneration. These findings reveal a self-reinforcing cycle wherein BMI decline reflects incipient pathology and amplifies disease progression through stage-specific mechanisms.

Key points: Question What is the association between changes in body mass index (BMI) and the pathological progression of Alzheimer's disease? Findings Early frontotemporal β-amyloid deposition predicts preclinical BMI decline, which in turn is associated with accelerated tau accumulation and neurodegeneration during symptomatic progression. Clinical relevance Monitoring BMI trajectories provides a low-cost approach to identifying individuals at high risk for Alzheimer's disease and tracking its pathological progression, highlighting the potential value of metabolic interventions during preclinical stages.

Mismatch imaging has become a key concept in neuroradiology, offering valuable insights into the pathophysiology of cerebrovascular and oncological conditions. By highlighting discrepancies between neuroimaging parameters, mismatch-based algorithms have revolutionized diagnosis, treatment planning, and patient prognosis. In stroke-related clinical scenarios, the mismatch concept is now essential in identifying candidates for thrombolysis or estimating the stroke onset time. However, the increasing use of mismatched terminology can lead to confusion, particularly when the exact mismatch target or required imaging modalities are unclear. Concerning stroke evaluation, there is a wide range of computed tomography perfusion (CT) maps and MRI sequences that are currently used for describing and determining mismatch concepts. Apart from the well-known penumbra-core mismatch related to CTP, the combination of features of different MRI sequences has provided a wide range of mismatch scenarios, such as perfusion-weighted imaging (PWI)/diffusion-weighted imaging (DWI) mismatch, magnetic resonance angiography (MRA)/DWI mismatch, susceptibility-weighted imaging (SWI)/DWI mismatch, or the DWI/FLAIR mismatch. Each one of these mismatches has its own clinical and physiopathological meaning, ranging from time-to-onset stroke estimation to selection of endovascular procedures. This article explores the different mismatch concepts used for stroke evaluation, including other related, less-used ones, focusing on their underlying physiopathology, clinical relevance, and supporting scientific evidence, all from a practical and educational perspective. KEY POINTS: Question Mismatch imaging offers transformative diagnostic insights in stroke, yet consistent definitions and standardized methodologies are essential to fully realize its clinical potential. Findings Our study reviews mismatch paradigms, outlines their pathophysiological basis, and compares clinical implications, highlighting critical limitations and the need for standardized imaging methodologies. Clinical relevance Standardizing mismatch imaging enhances diagnostic accuracy, optimizes patient selection for therapies, and improves prognostic assessment, ultimately enabling more consistent and reliable clinical decision-making in stroke.

Perfusion MRI techniques-including dynamic susceptibility contrast (DSC) MRI, dynamic contrast-enhanced (DCE) MRI, arterial spin labeling (ASL) MRI, and intravoxel incoherent motion (IVIM) MRI-hold strong potential as imaging techniques for diagnosing, staging, and monitoring disease across a range of clinical applications. However, clinical adoption, especially of quantitative parameters, remains variable across techniques. Key barriers to broader implementation include a lack of standardized acquisition and analysis protocols, leading to poor reproducibility and reduced clinical confidence. Additionally, limited awareness and understanding of certain techniques among radiologists contribute to underutilization in practice. This work provides practice recommendations to support radiologists in integrating perfusion MRI into routine clinical workflows. It includes guidance on technique selection, acquisition, and analysis, supported by a flowchart outlining typical imaging pathways. These efforts align with ongoing initiatives such as the Quantitative Medical Imaging Coalition (formerly QIBA) and the ISMRM Open Science Initiative for Perfusion Imaging (OSIPI), which are developing standards and tools to enhance reproducibility and clinical utility. Ultimately, the successful adoption of state-of-the-art perfusion MRI depends on close collaboration between clinicians, researchers, and industry stakeholders to ensure robust, standardized, and clinically meaningful application. KEY POINTS: Perfusion MRI parameters hold great promise as imaging biomarkers, but their clinical adoption, especially of quantitative parameters, remains variable across perfusion MRI techniques. An overview of perfusion MRI techniques, explaining the physics, illustrating clinical applications, and addressing common technical challenges, is provided to support perfusion MRI use in clinical practice. Successful adoption of state-of-the-art perfusion MRI depends on close collaboration between clinicians, researchers, and industry stakeholders to ensure robust, standardized, and clinically meaningful applications for patient care.

Objectives: Intracranial hemorrhage (ICH) is a time-critical neurological emergency in which rapid CT-based assessment directly informs treatment decisions. This study aimed to develop an automated deep-learning pipeline to enhance ICH detection, segmentation, and localization, complemented by clinical decision-making support through a large language model.

Materials and methods: The detection model was trained on 21,784 labeled and 3528 unlabeled CT scans from the RSNA dataset using semi-supervised learning. The segmentation model was trained on 1226 scans from the HS dataset to delineate six ICH subtypes. Hydrocephalus and midline-shift models were trained on a dedicated 507-scan subset of the HS dataset. Hemorrhage and edema locations were registered to standard brain regions to improve interpretability. For evaluation, the CQ500 dataset (491 patients) was used as an external validation and test cohort. Clinical recommendations were generated using the GPT-4o Assistants API based on published guidelines and trials.

Results: On the test set, detection achieved an AUC of 0.96 (95% CI: 0.94-0.98), and segmentation yielded Dice values ranging from 0.71 to 0.93 with corresponding 95% CIs from 0.61-0.76 to 0.90-0.96, while volume estimation showed high concordance (CCC 0.820-0.996). Intraparenchymal hemorrhage (IPH) localization demonstrated strong agreement with κ values of 0.85-1.00 across brain regions. Clinical decisions generated by the pipeline were highly rated, with one neurosurgeon assigning median scores of 4 and 5 for examination and treatment, and the other assigning 5 for both.

Conclusions: This deep learning pipeline combines imaging analysis with actionable clinical decisions, demonstrating significant potential as a valuable tool for emergency care.

Key points: Question Rapid and accurate identification of ICH on CT is critical for guiding treatment, yet remains difficult using standard emergency radiological evaluation. Findings The end-to-end artificial intelligence pipeline achieved high accuracy in ICH detection, segmentation, and localization, with strong concordance to manual measurements and reliable clinical recommendations. Clinical relevance By automating image analysis and clinical decision-making, the pipeline demonstrated significant potential to reduce diagnostic delays, improve treatment guidance, and enhance patient outcomes in emergency care settings.