This study aimed to develop a predictive model integrating clinical features and multisequence MRI radiomics to forecast postoperative seizure outcomes in pediatric patients with low-grade epilepsy-associated tumors (LEATs) who underwent gross total resection (GTR). In this study, we propose a novel radiomics-based approach to predict seizure recurrence. The model was further optimized by integrating clinical features, and its performance was compared with traditional radiomics models and deep learning-derived radiomics models. For traditional radiomics models, multi-sequence combination (Combined) outperformed single sequences, with XGBOOST achieving the highest AUC (0.889) and accuracy (0.816). Integrating preoperative epilepsy duration significantly improved model efficacy. The combined model of multimodal MRI radiomics and clinical features demonstrates potential for predicting postoperative seizure outcomes in pediatric LEAT patients after GTR.
Disease-modifying therapies for degenerative ataxias, including emerging gene therapies, are in the clinical trial pipeline. Sensitive outcome measures are urgently needed for treatment monitoring and participant selection to improve trial feasibility in these rare diseases. In hereditary ataxias, the ability to identify people who carry disease-causing mutations before symptom onset also raises the possibility of enrolment into trials before symptom manifestation, but biomarker data are key for participant selection and outcome monitoring in such preventive trials. Quantitative neuroimaging readouts have emerged as viable candidate biomarkers of ataxia pathology and progression that could supplement traditional clinical outcome assessments as clinical trial end-points. In this Consensus Statement, the Ataxia Global Initiative MRI Biomarkers Working Group critically reviews candidate MRI end-points for trials in the most common spinocerebellar ataxias (SCA1, SCA2 and SCA3) and Friedreich ataxia and provides evidence-based, disease-specific recommendations for the selection of MRI end-points for trials in these diseases. Recommendations are also provided for further research to address remaining knowledge gaps.
Supratentorial pediatric ependymomas (ST-EPNs) represent a rare and biologically heterogeneous subgroup of central nervous system tumors. The 2021 WHO molecular classification has redefined their diagnostic framework, emphasizing prognostic and therapeutic implications. This study analyzes clinical, surgical, and molecular features of pediatric ST-EPNs treated at a single-institution, with particular focus on outcomes and prognostic variables. We retrospectively reviewed pediatric patients with histologically confirmed supratentorial ependymomas treated between 2007 and 2023. Tumors were reclassified according to the 2021 WHO molecular criteria. Clinical presentation, imaging characteristics, tumor location, molecular subtype, extent of resection, adjuvant therapies, recurrence patterns, and long-term outcomes were evaluated. Twelve patients were included (mean age 7.4 years). Two tumors were intraventricular and ten extraventricular. Molecular profiling identified ZFTA fusion-positivity in 11 cases and YAP1 fusion-positivity in one case. Gross total resection was achieved in 11 of 12 patients, whereas one patient underwent near-total resection because of vascular encasement of the middle cerebral artery, making complete resection unsafe. Intraventricular tumors presented earlier, required complex surgical management, and were associated with higher treatment burden. Extraventricular tumors commonly presented with seizures and cystic imaging features, with favorable postoperative recovery. Recurrence occurred in 3 patients, successfully managed with repeat surgery and adjuvant therapy. At a mean follow-up of 8 years, all patients were alive with good neurological outcomes. Secondary neoplasms were observed in isolated cases following multimodal treatment. This single-institution molecularly reclassified series highlights clinically relevant differences between intraventricular and extraventricular pediatric supratentorial ependymomas in presentation, surgical complexity, and treatment burden. Our findings support maximal safe resection as the central therapeutic strategy while emphasizing the value of integrated molecular diagnostics and prolonged surveillance, particularly in light of late recurrence and secondary neoplastic events.
Carotid webs (CWs) are an underrecognized cause of cryptogenic ischemic stroke and transient ischemic attack (TIA). Morphological characteristics are believed to influence their potential to become symptomatic. We aimed to develop and internally validate a reliable predictive model for symptomatic CWs. We conducted an institutional retrospective study including patients with CWs identified on Computed Tomography Angiography (CTA). Demographic, clinical, and morphological data were extracted. Symptomatic CWs were defined by a history of stroke or TIA without a more likely cause based on TOAST criteria. To ensure robust internal validation and mitigate overfitting, 5-fold cross-validation was employed. Univariable logistic regression was applied to all variables, followed by multivariable analysis including predictors with p < 0.15. Model performance was assessed using the cross-validated area under the receiver operating characteristic (ROC) curve (AUC). A total of 124 CWs from 115 patients were identified, comprising 23 symptomatic and 101 asymptomatic webs (identified in 92 patients). The symptomatic group had higher rates of hypertension (91% vs. 59%, p = 0.003), hyperlipidemia (91% vs. 61%, p = 0.006), and CWs that were longer (3.0 mm [IQR 2.5-3.6] vs. 2.3 mm [IQR 2.0-3.0], p = 0.002) and thicker (1.5 mm [IQR 1.0-2.8] vs. 1.0 mm [IQR 0.9-1.5], p = 0.008). In the multivariable model, female sex (OR: 4.11, 95% CI: 1.31-15.3, p = 0.022), hypertension (OR: 7.58, 95% CI: 1.86-52.4, p = 0.013), and stenosis (OR: 1.06, 95% CI: 1.02-1.11, p = 0.003) were independent predictors of symptomatic status. The model yielded a 5-fold cross-validated AUC of 0.78, with a sensitivity of 91% and a specificity of 60% at the optimal probability threshold. Female sex, hypertension, and local stenosis are independently associated with symptomatic CWs. This cross-validated model provides a framework for identifying high-risk carotid webs and may aid in clinical risk stratification.
Polymorphous low-grade neuroepithelial tumor of the young (PLNTY) was recognized as a new histologic entity in the 2021 World Health Organization (WHO) classification of central nervous system tumors. It represents a distinct epileptogenic tumor, with about 80 cases reported to date across all age groups. The aim of this study is to describe the features of a cohort of PLNTY patients with a multimodal approach. Patients diagnosed with PLNTY from 2014 to 2023 at our Institute were retrospectively reviewed to collect clinical, EEG, neuroimaging, histopathological, and molecular data. We identified 14 surgically treated patients (median age 21.5 years, range 12-46), with drug-resistant seizures in 78.6%. Seizure onset preceded surgery by a median period of 4 years (range .5-27). Bilateral interictal epileptiform EEG discharges and/or mislateralizing/mislocalizing ictal EEG discharge and/or semiological signs were found in 85.7% of patients, compared to 19.2% in a group of 26 non-PLNTY low-grade epilepsy-associated tumors (LEATs) (p < .001). All PLNTYs were supratentorial solid or solid-cystic cortical mass, involving the temporal (n = 8), temporo-occipital (n = 4), frontal (n = 1), or parietal lobe (n = 1). They featured an infiltrative growth pattern, oligodendroglioma-like cells, strong and often diffuse CD34 immunostaining, and frequent calcifications. Focal cortical dysplasia was associated in 14.3% of cases. Proliferation activity (MIB-1 labeling index) was very low (1%), except for one case (3%). BRAFV600E was found in 50% (n = 6/12) of cases, fibroblast growth factor receptor 3 (FGFR3) protein expression in 7 cases, 30.7% (n = 4/13) carried FGFR3 fusion. No fibroblast growth factor receptor 2 (FGFR2) fusion (n = 0/6) was identified. Postsurgical seizure outcome was excellent (Engel class I) in all cases. This study confirms that, despite its name, PLNTY is not limited to pediatric patients. Findings underscore the highly epileptogenic nature of PLNTY and its recognizable electroclinical features, potentially related to its distinctive neuropathology. Most PLNTYs show mitogen-activated protein kinase (MAPK) pathway activating alterations, demonstrated by BRAFV600E mutation and FGFR3 fusion.
Cerebral malaria (CM) impacts hundreds of thousands of African children and causes significant morbidity and mortality. Beyond supportive care and antimalarials, treatments that improve outcomes have not been identified. We conducted a randomized, open-label, phase 3 clinical trial in Blantyre, Malawi (2018-2023). Children aged 6 months to 12 years with CM (Blantyre Coma Score ≤ 2, Plasmodium falciparum parasitemia, and no alternative coma etiology) underwent magnetic resonance imaging (MRI) screening. Participants with severe brain swelling [brain volume score (BVS) 6-8] were randomized to early intubation and mechanical ventilation, 3% hypertonic saline (HTS), or usual care. The primary outcome was 7-day mortality. Secondary outcomes included neurologic morbidity in survivors, serious adverse events, and changes in transcranial Doppler (TCD)-measured cerebral blood flow velocities (CBFVs). Of 206 children screened with MRI, 57 met criteria for randomization (23 mechanical ventilation, 17 HTS, and 17 usual care). Neither mechanical ventilation nor HTS reduced 7-day mortality compared with usual care (risk difference - 0.075, 95% CI - 0.27 to 0.17, p = 0.59 for ventilation; 0.002, 95% CI - 0.22 to 0.28, p = 1.0 for HTS). Neurologic morbidity at hospital discharge did not differ between intervention groups and controls. HTS administration resulted in rapid and sustained increases in CBFVs in all major cerebral vessels. Among children with baseline cerebral hyperemia, those treated with HTS had significantly worse outcomes (combined neurologic morbidity or death) than hyperemic children receiving usual care (relative risk 3.0, 95% CI 1.02 to 10.9, p = 0.04). In children with CM and severe brain swelling, adjunctive early mechanical ventilation or hypertonic saline does not improve survival or neurologic outcomes. Hypertonic saline increases cerebral blood flow and is associated with harm in children with baseline cerebral hyperemia. These findings demonstrate that non-phenotype-guided neuroprotective strategies are ineffective in CM and highlight the need for cerebrovascular endotype-informed therapeutic approaches.
Granulomatous disorders of the central nervous system (CNS) encompass a highly heterogeneous group of diseases with autoinflammatory, autoimmune, infectious, proliferative-neoplastic, foreign body-related, or drug-induced aetiologies. Neuroimaging plays a pivotal role, particularly with respect to early diagnosis. Therefore, knowledge of specific imaging patterns is crucial in order to narrow down the broad differential diagnosis. Although imaging cannot replace histology, its importance in initiating therapy as early as possible is indisputable. This is particularly true since many granulomatous diseases respond to treatment, usually with immunosuppressants and/or anti-infectives. This review highlights representative disease entities across the major aetiological categories and discusses key neuroradiological features and clinical symptoms together with relevant differential diagnostic considerations.
暂无摘要（点击查看详情）
Prolonged acquisition times and motion artifacts limit the use of MRI in the pediatric emergency department, frequently necessitating sedation and reducing workflow efficiency, or prompting the use of CT with associated ionizing radiation. Recent MRI advances enable fast, comprehensive neuroimaging without radiation exposure. This study evaluates the diagnostic interchangeability of a fast 2.5-minute MRI multicontrast sequence (NeuroMix) with a standard MRI protocol for detecting clinically relevant intracranial findings in the pediatric emergency department. We retrospectively identified sequential pediatric patients who underwent both a standard pediatric emergency department MRI protocol and NeuroMix over a one-year period. A finding-enriched cohort of 100 studies was constructed (44 positives and 56 controls). Three radiologists independently reviewed NeuroMix and standard protocol imaging sets in randomized batches. Diagnostic interchangeability was assessed using an agreement-based mixed-effects modeling approach using a noninferiority threshold of 10%. Acquisition times, examination completeness, artifact prevalence, interreader agreement, and diagnostic performance metrics were compared between protocols, in accordance with STARD. NeuroMix demonstrated diagnostic interchangeability with the standard pediatric emergency MRI protocol for detection of clinically relevant findings, with an interchangeability estimate of -3.1% (95% CI, -6.6% to +0.3%), remaining within the noninferiority margin. Overall diagnostic accuracy was identical for NeuroMix and standard imaging (87.7%). Interreader agreement for detection of clinically relevant findings was substantial for both NeuroMix (κ = 0.65) and the standard protocol (κ = 0.73). Mean acquisition time was markedly shorter for NeuroMix when compared with the standard protocol (median 2.67 minutes vs 15.26 minutes, p < 0.001). Moderate to severe artifacts and repeat sequence acquisitions occurred less frequently with NeuroMix. A 2.5-minute NeuroMix-based MRI protocol demonstrated diagnostic interchangeability with a standard pediatric emergency MRI protocol for clinically relevant intracranial findings with substantially reduced acquisition time and imaging artifacts. These results support NeuroMix as a fast, radiation-free approach that may serve, depending on the clinical scenario, as a supplementary or alternative neuroimaging protocol in pediatric emergency care.
The Union for International Cancer Control (UICC)/American Joint Committee on Cancer (AJCC) 9th version Tumor-Nodes-Metastasis (TNM) staging incorporates iENE to improve prognostic accuracy in human papilloma virus-associated (HPV) oropharyngeal carcinoma. We evaluated the impact of iENE on reclassification and associated clinical outcomes in HPV oropharyngeal carcinoma in a retrospective study. Patients with HPV oropharyngeal carcinoma with positive lymph-nodes at diagnosis and treated with curative-intent chemoradiotherapy were included. Pre-treatment MRI images were reviewed by a head and neck radiologist to assess for iENE. All patients were (re)staged according to the 8th and the 9th TNM versions. Disease-free survival (DFS) was compared between cases with and without iENE and across clinical nodal categories. Twenty-six consecutive patients with HPV oropharyngeal carcinoma with positive nodes were analyzed. All patients completed the planned treatment and achieved a clinical complete response at 3 months post-therapy. iENE was identified in 16 patients (61.5%). The presence of iENE was associated with lower 2-year DFS: 68.8% (95% CI 0.41-0.86) versus 100% (95% CI not estimable) (p = 0.048). Reclassification using the 9th TNM edition improved prognostic discrimination compared with the 8th edition (p = 0.043). The 2-year DFS rates by 9th edition clinical nodal stage were 100% (95% CI not estimable), 75% (95% CI 0.41-0.91) and 66.7% (95% CI 0.19-0.90) for clinical nodal 1, 2 and 3, respectively, compared to 82.9% (95% CI 0.56-0.94), 71.4% (95% CI 0.26-0.92) and 100% (calculated at 12 months) for the 8th edition. The 9th TNM version seems to improve DFS stratification in HPV oropharyngeal carcinoma after definitive chemoradiotherapy.
Giant cell arteritis (GCA) is the most common form of vasculitis in people over the age of 50. It constitutes a true medical emergency due to the risk of blindness or other ischemic manifestations. Treatment should therefore ideally begin within hours of the onset of the first clinical signs suggestive of this diagnosis. However, the often complex and nonspecific clinical presentation, combined with the absence of clear diagnostic criteria, frequently leads to delays in management. We report two cases of patients with giant cell arteritis presenting with very distinct manifestations. The first case is a classic history of Horton's disease, rapidly diagnosed due to the typical clinical presentation. In the second case, the diagnosis required additional imaging studies, notably a PET-CT scan, due to the nonspecific yet common presentation of giant cell arteritis, which should not be overlooked. Through these two cases, we aim to raise awareness among primary care clinicians regarding the various clinical presentations that should raise suspicion of giant cell arteritis and prompt the need for further investigations.
Microplastics (MPs) have been detected in human tissues and implicated in vascular injury in preclinical models, but evidence on circulating MPs in humans is scarce. Surgical and autopsy samples have provided most data to date but are unsuitable for large-scale biomonitoring to clarify potential cardiovascular relevance. We conducted a prospective pilot study applying a contamination-controlled, plastic-free protocol to quantify MPs in whole blood. Thirty-five participants were enrolled: 20 with ischemic stroke and 15 healthy controls. Samples were analyzed by micro-Fourier transform infrared spectroscopy to determine particle counts, polymer types, and size distributions. To contextualize findings, we performed an exploratory pooled comparison with a reference of healthy individuals derived from two published datasets using weighted combination of reported means and standard deviations. Overall, MPs were detected in 94% of participants (median 11.0 MPs/mL; IQR 4-22; mean ± SD, 12.6 ± 9.7; range 0-35). Within this pilot cohort, median MP counts were numerically higher in stroke (13.5 MPs/mL [IQR 6-23]) than in healthy controls (7.0 MPs/mL [IQR 2-18]), but group differences were not statistically significant in direct, within-study comparisons (p = 0.21, Mann-Whitney U). In an exploratory analysis, stroke patients exhibited significantly higher mean MP concentrations than a pooled healthy reference derived from external datasets (13.95 ± 8.73 vs 5.11 ± 7.31 MPs/mL; Welch's t = 3.69, p = 0.0016). Polyethylene, polypropylene, and polyethylene terephthalate were the most abundant polymers, and smaller particles (20-100 µm) predominated. MP burden showed no consistent association with age, sex, or vascular risk factors. This proof-of-concept study demonstrates the feasibility of blood-based MP quantification under clinical conditions and provides a preliminary, hypothesis-generating observation that circulating MP levels may be elevated in acute ischemic stroke. Because the apparent difference relied on an exploratory pooled comparison across heterogeneous datasets, confirmation in larger, harmonized studies is essential before any inference about disease-specific elevation or clinical relevance is justified.
Acute silent ischemic lesions (ASILs) usually present as diffusion weighted imaging hyperintensity with corresponding apparent diffusion coefficient hypointensity, indicating a possibly unstable ischemic pattern in severe intracranial stenosis. However, there is still no conclusive evidence that preprocedural ASILs are associated with inhospital stroke after percutaneous transluminal angioplasty and stenting (PTAS). A single center retrospective cohort study was conducted among patients with severe symptomatic intracranial atherosclerotic stenosis (severe sICAS) who underwent PTAS at a tertiary care center between January 2019 and October 2024. Among 1727 initially screened patients, 1427 eligible patients were included according to preprocedural ASIL status. The primary endpoint was prespecified as stroke or death during the index hospitalization. Multivariable logistic regression was used to evaluate the association between ASILs and the primary endpoint. Sensitivity analyses included propensity score matching, inverse probability of treatment weighting using winsorized stabilized weights, doubly robust weighted modeling using the same winsorized stabilized weights, Firth penalized logistic regression, early period exclusion analyses, and parsimonious models addressing events-per-variable concerns. Apparent within cohort discrimination and reclassification were assessed by the area under the receiver operating characteristic curve, DeLong test, continuous net reclassification improvement, and integrated discrimination improvement. Of the 1427 participants, 60 (4.2%) had the primary endpoint, all of which were inhospital stroke events; no inhospital deaths occurred. Preprocedural ASILs were associated with inhospital stroke in univariable analysis (OR 4.62, 95% CI 2.68 to 7.98; P<0.001) and remained independently associated with inhospital stroke in the prespecified multivariable model (OR 5.00, 95% CI 2.70 to 9.26; P<0.001). The association remained consistent in propensity score matching (OR 4.20, 95% CI 1.58 to 11.14; P=0.004), inverse probability of treatment weighting using winsorized stabilized weights (OR 3.94, 95% CI 2.16 to 7.18; P<0.001), doubly robust weighted modeling using the same winsorized stabilized weights (OR 4.12, 95% CI 2.26 to 7.52; P<0.001), and Firth penalized logistic regression (OR 4.72, 95% CI 2.57 to 8.56; P<0.001). Adding ASIL status to the prespecified clinical model increased the apparent AUC from 0.705 to 0.763 (DeLong P=0.010) and improved reclassification. Preprocedural ASILs were independently associated with inhospital stroke after PTAS in patients with severe sICAS. Incorporation of ASIL status into a prespecified clinical model was associated with improved apparent within cohort discrimination and reclassification, but external validation is required before clinical application.
Intracerebral haemorrhage (ICH) accounts for approximately 15% of all strokes in Denmark and remains associated with high mortality and morbidity. It is challenging to distinguish neoplastic from non-neoplastic causes of ICH in the acute setting, and CT findings that may aid early differentiation have not been fully characterised. Existing ICH-classification systems (SMASH-U, H-ATOMIC and CLAS-ICH) have not been directly compared for diagnostic accuracy in this setting. Identifying radiological and clinical factors associated with underlying aetiology may support faster diagnosis, reduce time to workup related to potential underlying cancer and facilitate early targeted treatment of the underlying cause of ICH. This study is a retrospective observational cohort including all patients admitted with acute ICH to the Department of Neurology, University Hospital of Southern Denmark, Aabenraa between January 2014 and December 2024 (estimated approximately n=610). Medical records and initial non-enhanced CT scans will be reviewed. Two neurologists and two radiologists, blinded to final diagnosis, will independently extract clinical presentation, topographical and volumetric haemorrhage characteristics, and classify each case using the abovementioned ICH-classification systems. Primary analyses will assess associations between clinical and radiological features and underlying neoplastic vs non-neoplastic aetiology. Secondary analyses will compare diagnostic performance of classification systems using sensitivity, specificity and receiver operating characteristic curves. Multivariate logistic regression models will be applied with Holm correction for multiple comparisons. The study has been submitted to the National Danish Research Ethics Committee and the Danish Data Protection Agency. As data derive from completed disease courses, no patient contact is expected. Results will be disseminated through peer-reviewed journals, conferences and scientific presentations.
Molecular subtyping guides diagnosis and targeted therapy for gliomas. Although MRI-the current imaging standard-can be time-consuming and is sometimes contraindicated, computed tomography (CT) is faster, more widely available, and often preferable in emergency and resource-limited settings. We evaluated whether CT-based radiogenomic signatures combined with machine learning could accurately predict clinically relevant glioma molecular markers. In this retrospective study, we characterised non-contrast CT (NCCT) scans from 197 adults with histologically confirmed gliomas. Models were developed to predict mutations in four molecular markers: ATRX (n=81), EGFR (n=17), TP53 (n=71), and IDH (n=183). We extracted 208 quantitative radiomic features and added basic demographic variables. Feature selection used LASSO-RFE and Gradient Boosting-RFE with cross-validation. Six classical machine-learning classifiers and deep-learning approaches - including custom fully connected neural networks (FCNN) and TabNet - were trained and compared using ROC-AUC as the primary performance metric. Deep-learning methods outperformed conventional classifiers for all targets. TabNet achieved ROC-AUCs of 0.900 (95% CI: 0.717-0.989; ATRX), 0.955 (95% CI: 0.661-0.978; TP53), and 0.917 (95% CI: 0.858-0.975; EGFR). A custom FCNN obtained a ROC-AUC of 0.971 (95% CI: 0.876-0.995) for IDH. Cross-validation coefficients of variation were 2.0% for ATRX and 3.0% for TP53 and EGFR, and 16.0% for IDH. Deep-learning approaches yielded statistically significant improvements over conventional methods (p-values ranging from <0.05 to <0.001). NCCT-based analytical methods were able to predict clinically relevant genetic mutations in gliomas and demonstrated performance comparable to established techniques. These findings suggest that CT may serve as a practical option for molecular profiling in urgent or resource-limited settings. Nonetheless, external validation is necessary prior to clinical translation. The EGFR findings, arising from a small EGFR-tested subgroup (n = 17), remain preliminary; the observed performance may not reflect generalisable accuracy, and independent validation in larger cohorts is required.
Digital subtraction angiography (DSA) interpretation is observer dependent. This study evaluated the diagnostic performance of an existing deep-learning (DL) based thrombus classifier prior to clinical application. The intended use of the model is as a clinical decision-support tool to assist neuroradiological assessment during mechanical recanalization. This retrospective study included an in-house dataset of DSA image series from endovascular recanalization procedures for anterior circulation occlusions collected over two years. For each case, two DSA runs were selected: one before and one after recanalization. The artificial intelligence system was applied to classify thrombus presence. Diagnostic performance was assessed using sensitivity, specificity, and false-positive rate. A total of 1,236 DSA series from 309 patients were analyzed, yielding 618 paired biplane acquisitions. The DL classifier achieved an overall sensitivity of 71.7% (95% CI 67.3-75.8%), with the highest sensitivity for proximal vessel occlusions (M1/M2 segments: 87.6%; 95% CI 83.6-90.9%) and substantially lower sensitivity for distal occlusions (M3/M4 segments: 23.1%; 95% CI 14.9-33.1%) as well as for occlusions of the anterior cerebral artery (27.3%; 95% CI 10.7-50.2%). Overall specificity for thrombus detection was 89.8% (149/166) (95% CI: 84.1-93.9%), corresponding to 17 false-positive classifications. The developed DL classifier on DSA series confirmed on a large-scale dataset its very high sensitivity to proximal vessel occlusions with a sensitivity of 87.6%. Sensitivity for distal vessel occlusions was very low. Training the system for these lesion types will be the next step prior to clinical application.
Peptide receptor radionuclide therapy (PRRT) with radiolabelled somatostatin analogues, such as [¹⁷⁷Lu]Lu-DOTATATE, is approved as a second-line treatment for progressive, gastroenteropancreatic neuroendocrine tumors (GEP-NETs). However, emerging evidence suggests potential earlier roles for PRRT. A systematic review was conducted in accordance with the PRISMA guidelines, including case reports, original papers, and clinical trial protocols evaluating neoadjuvant, adjuvant, or first-line PRRT in GEP-NETs. A systematic search of PubMed and clinicaltrials.gov (up to May 31, 2025) finally identified 25 eligible papers. For each paper, data were extracted regarding clinical setting, patient characteristics, treatment regimen, and oncologic and surgical outcomes. The available preliminary data, comprising case reports, retrospective and prospective cohorts, show that neoadjuvant PRRT can induce tumor shrinkage, regression of vascular involvement, and conversion from unresectable to resectable disease in up to 45% of patients. Reported outcomes include high objective response rates (up to 70%), improved surgical feasibility, and a promising survival benefit, especially in patients who received PRRT before surgery. Also, preliminary results from the NETTER-2 trial support the use of [¹⁷⁷Lu]Lu-DOTATATE as first-line therapy in SSTR-positive G2-G3 GEP-NET, showing significant improvement in progression-free survival. Differently, the evidence for adjuvant use is missing. Current data suggest that PRRT, traditionally approved as a later-line therapy, may play a promising role in neoadjuvant and first-line settings for selected GEP-NET patients, enhancing resectability, improving surgical outcomes, and potentially prolonging survival. Further prospective, randomized studies are needed to define selection criteria, determine optimal timing, and assess the long-term impact on disease trajectory.
The practice parameter was revised collaboratively by the American College of Radiology (ACR), the American Brachytherapy Society (ABS), the American College of Nuclear Medicine (ACNM), the American Radium Society (ARS), the Society of Interventional Radiology (SIR), and the Society of Nuclear Medicine and Molecular Imaging (SNMMI). Epidural steroid injections (ESI) are widely used procedures for managing spinal pain and radiculopathy in patients who do not respond to conservative therapy. This multidisciplinary practice parameter reviews evidence for the efficacy and safety of ESI and provides best practice recommendations. This practice parameter was revised according to the process described under the Process for Developing ACR Practice Parameters and Technical Standards on the ACR website (https://www.acr.org/Clinical-Resources/Clinical-Tools-and-Reference/Practice-Parameters-and-Technical-Standards) by the Committee on Practice Parameters - Interventional and Cardiovascular Radiology of the ACR Commission on Interventional and Cardiovascular, Committee on Practice Parameters and Technical Standards - Nuclear Medicine and Molecular Imaging of the ACR Commission on Nuclear Medicine and Molecular Imaging and the Committee on Practice Parameters - Radiation Oncology of the ACR Commission on Radiation Oncology in collaboration with the ABS, the ACNM, the ARS, the SIR, and the SNMMI. Epidural steroid injections can provide meaningful improvements in pain, mobility, function, and quality of life for appropriately selected patients with neck and low back pain, particularly when degenerative changes or radiculopathy are present. This document advises that ESI be performed with image guidance, meticulous technique, and detailed knowledge of spinal anatomy, particularly in light of rare but serious neurologic risks highlighted by the FDA. The parameter specifies absolute and relative contraindications, encourages integrating ESI within a multidisciplinary care pathway rather than as stand-alone therapy, and recommends that only appropriately trained and credentialed physicians who understand the indications, relevant pharmacology, and complication management perform these procedures. This multidisciplinary practice parameter provides recommendations on indications, contraindications, patient selection, and technical execution to optimize the therapeutic benefit of epidural steroid injections while reducing the risk of serious neurologic and other complications.
Artificial intelligence is increasingly integrated in clinical practice. In radiological imaging, deep-learning (DL)-based image reconstruction techniques show potential for accelerating and enhancing the quality of examination procedures in magnetic resonance imaging (MRI). This study evaluated the impact of DL on MRI workflows and protocols over a one-year period following clinical implementation. This retrospective, single-center study included 8,183 MRI examinations performed between 2023 and 2024 to assess daily examination volume, examination duration, and rates of repeated and aborted scans. Furthermore, 43 comparable sequences and 34 protocols on a 1.5T MRI (Siemens Magnetom Sola) were analyzed and surveys were conducted among 23 medical staff members to evaluate the perceived effects of DL on workflow efficiency, diagnostic performance, stress levels, and technology acceptance. 88% of the DL-reconstructed sequences demonstrated reduced acquisition times, yielding an overall 13% reduction in protocol duration, despite higher resolution or reduced slice thickness. On days without anesthesia support, the average examination time decreased by 11%, accompanied by increased patient throughput (+7.2%) and fewer repeats (-25%) and aborts (-100%). Medical staff reported high levels of technology acceptance (90%), perceived improvements in image quality (90.5%), and reduced stress levels. Shorter times required for generating medical reports were noted by 45.5% of respondents, particularly among residents (70%). DL enables significant potential for MRI-workflow optimization and protocol improvements, while maintaining high staff satisfaction, thereby highlighting the great potential of DL in radiology practice. · While maintaining or increasing image quality, DL reduced examination duration (-11%) and protocol length (-13%).. · Patient throughput increased (+7.2%), while repeats (-25%) decreased.. · Staff reported high technology acceptance, lower stress and faster report generation, particularly residents.. · Rizzetti M, Michael AE, Almansour H et al. Impact of Deep-Learning Reconstruction on MRI Workflows: A Retrospective Analysis at a Large Academic Tertiary Center. Rofo 2026; DOI 10.1055/a-2857-0974. Künstliche Intelligenz wird zunehmend in die klinische Routine integriert. Insbesondere Deep-Learning (DL)-basierte Rekonstruktionsverfahren zeigen in der Magnetresonanztomografie (MRT) ein hohes Potenzial zur Beschleunigung von Untersuchungsabläufen und Verbesserung der Bildqualität. Ziel dieser Studie war es, den Einfluss DL-gestützter Bildrekonstruktion auf MRT-Arbeitsabläufe und Untersuchungsprotokolle ein Jahr nach klinischer Implementierung zu evaluieren.In dieser retrospektiven monozentrischen Studie wurden 8183 MRT-Untersuchungen aus den Jahren 2023–2024 hinsichtlich Untersuchungsdauer, Patientendurchsatz und Häufigkeit wiederholter und abgebrochener Untersuchungen analysiert. Zusätzlich wurden 43 vergleichbare Sequenzen und 34 Protokolle an einem 1,5-T-Scanner (Siemens Magnetom Sola), ausgewertet und eine Befragung von 23 medizinischen Mitarbeitenden zur Einschätzung von Workflow-Effizienz, Bildqualität, Stressbelastung und Technologieakzeptanz durchgeführt.Bei 88% der DL-rekonstruierten Sequenzen verkürzte sich die Akquisitionszeit, was trotz höherer räumlicher Auflösung oder reduzierter Schichtdicke zu einer durchschnittlichen Reduktion der Protokolldauer um 13% führte. An Tagen ohne anästhesiologische Unterstützung verringerte sich die mittlere Untersuchungsdauer um 11%, bei gleichzeitig erhöhtem Patientendurchsatz (+7,2%) sowie deutlich reduzierten Wiederholungen (–25%) und Abbrüchen (–100%). Das Personal berichtete über eine hohe Technologieakzeptanz (90%), wahrgenommene Verbesserungen der Bildqualität (90,5%) sowie reduzierte Stressbelastung. Verkürzte Befundungszeiten wurden von 45,5% der Befragten angegeben, insbesondere von Assistenzärzten (70%).DL ermöglicht eine relevante Optimierung von MRT-Arbeitsabläufen und Untersuchungsprotokollen bei hoher Nutzerzufriedenheit und unterstreicht das Potenzial von DL für die radiologische Routine. · Bei gleichbleibender oder verbesserter Bildqualität reduzierte DL die Untersuchungsdauer (−11%) sowie die Protokolllänge (−13%).. · Der Patientendurchsatz stieg (+7,2%), während Wiederholungen (-25%) und Abbrüche (-100%) sanken.. · Das Personal berichtete über hohe Technologieakzeptanz, geringere Stressbelastung und schnellere Befundung, insbesondere bei Assistenzärzten..
Cluster headache (CH) is a highly disabling primary headache disorder increasingly conceptualized as a disorder of distributed brain networks rather than an isolated trigeminovascular dysfunction. While structural and functional MRI studies have demonstrated persistent brain alterations outside the ictal phase, the role of white matter microstructural changes supporting network-level dysfunction remains incompletely understood. We investigated cerebral white matter microstructure in 27 patients with episodic CH examined during the bout outside active attacks using diffusion tensor imaging and tract-based spatial statistics (TBSS). Fractional anisotropy (FA), axial diffusivity (AD), mean diffusivity (MD), and radial diffusivity (RD) were compared between patients and healthy controls (N = 20) on a whole-brain white matter skeleton. Associations between diffusion metrics and clinical variables were explored using an ANCOVA-based TBSS framework. Compared with controls, patients with episodic CH showed widespread white matter microstructural alterations, characterized by reduced FA and AD and increased MD in multiple clusters involving long-range associative and projection pathways connecting frontal, parietal, temporal, subcortical, and cerebellar regions. No significant associations were observed between diffusion metrics and clinical variables, including disease duration, attack frequency, and pain intensity. White matter alterations predominantly affected pathways supporting large-scale pain-related and cognitive-affective networks, including fronto-parietal, thalamo-cortical, and cerebello-thalamo-cortical connections. These findings are consistent with a network-based model of CH and integrate coherently with recent multimodal MRI evidence of cortical thickness and functional connectivity alterations observed outside attacks in a largely overlapping cohort. Our results provide evidence of persistent white matter microstructural abnormalities in episodic CH outside the ictal phase, supporting the concept of CH as a disorder of distributed brain networks. White matter alterations may represent a structural substrate underlying enduring network dysfunction predisposing patients to recurrent attacks.