Clinical Neuroradiology最新文献

Purpose: Techniques for endovascular thrombectomy in acute ischemic stroke due to medium or distal vessel occlusions (MDVO) need to be improved, and first-line aspiration is emerging as a promising technique. The objective was to describe first-pass and final reperfusion rates with a triaxial direct aspiration first pass technique (ADAPT) and a novel, quadriaxial technique (QUATTRO-ADAPT) using RED43 Reperfusion Catheters to target MDVOs.

Methods: This retrospective study collected data from 11 stroke centers in Europe and North America between 2023 and 2025. Patients with primary, isolated MDVOs who underwent thrombectomy using first-line aspiration with RED43 Reperfusion Catheters were included. Primary outcome was first-pass successful reperfusion rate (modified Treatment In Cerebral Infarction score ≥ 2b). Additionally, final reperfusion rates, complications and functional outcomes up to 90 days were evaluated.

Results: We included 85 cases. Median age was 74 years (IQR 67-81) and median National Institutes of Health Stroke Scale score 9 (IQR 6-15). The majority (69.4%) had an occlusion of the co- or nondominant M2 segment of the middle cerebral artery. First-pass successful reperfusion was 63.5% overall, with 55.6% for the triaxial ADAPT and 69.4% for the quadriaxial QUATTRO-ADAPT. Final successful reperfusion was achieved in 96.5%. Complications included vessel perforations (2.4%), subarachnoid hemorrhage (3.5%), parenchymal hematoma type 2 (2.4%), and symptomatic intracerebral hemorrhage (1.2%). Functional independence (modified Rankin Scale 0-2) at 90 days was observed in 58.7%.

Conclusion: This study shows that first-line aspiration, especially in a quadriaxial approach, can be effective and safe. These findings should be evaluated in prospective trials.

Background: Balloon-guide catheters (BGC) have been associated with improved procedural and functional outcomes following endovascular thrombectomy (EVT) in multiple observational studies. Recently the PROTECT-MT trial challenged this benefit, showing worse functional outcomes when using BGCs. This study aims to assess the association between BGC-use and procedural and functional outcomes in a large real-world cohort.

Methods: Patients who underwent EVT for anterior circulation large vessel occlusion (LVO) between 2017 and 2021 were included in two Swedish registries: EVAS and Riksstroke. Outcomes included recanalization success (modified Treatment In Cerebral Infarction [mTICI] 2b/2c-3), good 90-day functional outcome (modified Rankin Scale score 0-2), and safety outcomes (iatrogenic dissection, perioperative embolization, early neurological deterioration, and 90-day mortality). Subgroup analysis was conducted by first-line EVT strategy: stent-retriever (with/without aspiration) or contact aspiration alone.

Results: Of 4843 patients, 2483 met the inclusion criteria. BGCs were used in 1449 cases (58.4%) and were more frequently used with stent-retrievers (83.0%) than with contact aspiration (24.1%, p < 0.001). BGC-use was associated with higher rates of excellent recanalization (mTICI 2c-3) and first-pass success, particularly in stent-retriever cases (p < 0.001), but not with good recanalization (mTICI 2b-3). No association was found with good 90-day functional outcome or any safety outcome (p > 0.05). A numerically lower mortality rate was observed with BGCs (19.0% vs. 22.9%, p = 0.019), although this was not significant after adjustment.

Conclusions: BGC-use during EVT was associated with excellent recanalization and first-pass reperfusion, primarily in stent-retriever-treated patients. No significant impact on 90-day functional outcome nor on safety indicators was observed.

Purpose: Prognosis and therapeutic approaches to meningiomas are determined by their proliferative activity and tumor grade. The early identification of aggressive phenotypes is hence essential to improving outcomes. Here, we investigated the potential of shape-based radiomics features and surface regularity (SR) as non-invasive biomarkers of atypical (grade 2) and anaplastic (grade 3) meningiomas.

Methods: In this retrospective single-center study, we assessed individuals with treatment-naive meningiomas in a comprehensive analysis of MR imaging with histopathological grading and quantification of mitotic activity. Radiomics (sphericity, elongation, flatness) and SR measures were calculated on segmented contrast-enhancing tumor components and tested for association with WHO grade (n = 62 grade 1, n = 71 grade 2, and n = 19 grade 3) and proliferation.

Results: All tumor grades were significantly different in characteristics of sphericity and SR with decreasing values at higher tumor grades as an expression of a more irregular shape. Additionally, radiomics features of tumor elongation and flatness discriminated grade 3 cases from grade 1 (p = 0.0008 and p = 0.0003) and grade 2 meningiomas (p = 0.0008 and p = 0.0013). Mitotic activity exhibited a significant negative correlation with the four imaging markers assessed (p < 0.003).

Conclusion: Shape-based radiomics features of sphericity and SR can serve as preoperative grading biomarkers of meningiomas in routine contrast-enhanced T1w MRI.

Background: Non-contrast computed tomography (NCCT) is a first-line imaging technique for determining treatment options for acute ischemic stroke (AIS). However, its poor contrast and signal-to-noise ratio limit the diagnosis accuracy for radiologists, and automated AIS lesion segmentation using NCCT also remains a challenge. This study aims to develop a segmentation method for ischemic lesions in NCCT scans, combining symmetry-based principles with the nnUNet segmentation model.

Methods: Our novel approach integrates a Generative Module (GM) utilizing 2.5 D ResUNet and an Upstream Segmentation Module (UM) with additional inputs and constraints under the 3D nnUNet segmentation model, utilizing symmetry-based learning to enhance the identification and segmentation of ischemic regions. We utilized the publicly accessible AISD dataset for our experiments. This dataset contains 397 NCCT scans of acute ischemic stroke taken within 24 h of the onset of symptoms. Our method was trained and validated using 345 scans, while the remaining 52 scans were used for internal testing. Additionally, we included 60 positive cases (External Set 1) with segmentation labels obtained from our hospital for external validation of the segmentation task. External Set 2 was employed to evaluate the model's sensitivity and specificity in case-dimensional classification, further assessing its clinical performance. We introduced innovative features such as an intensity-based lesion probability (ILP) function and specific input channels for suspected lesion areas to augment the model's sensitivity and specificity.

Results: The methodology demonstrated commendable segmentation efficacy, attaining a Dice Similarity Coefficient (DSC) of 0.6720 and a Hausdorff Distance (HD95) of 35.28 on the internal test dataset. Similarly, on the external test dataset, the method yielded satisfactory segmentation outcomes, with a DSC of 0.4891 and an HD 95 of 46.06. These metrics reflect a substantial overlap with expert-drawn boundaries and demonstrate the model's potential for reliable clinical application. In terms of classification performance, the method achieved an Area Under the Curve (AUC) of 0.991 on the external test set, surpassing the performance of nnUNet, which recorded an AUC of 0.947.

Conclusion: This study introduces a novel segmentation technique for ischemic lesions in NCCT scans, leveraging symmetry-based principles integrated with nnUNet, which shows potential for improving clinical decision-making in stroke care.

Purpose: To investigate the feasibility of Retrieval-augmented Generation (RAG)-enhanced Large Language Models (LLMs) in answering questions about two German neurovascular guidelines.

Methods: Four LLMs (GPT-4o-mini, Llama 3.1 405B Instruct Turbo, Mixtral 8 × 22B Instruct, and Claude 3.5 Sonnet) with RAG as well as GPT-4o-mini without RAG were evaluated for generating answers about two German neurovascular guidelines ("S3 Guideline for Diagnosis, Treatment, and Follow-up of Extracranial Carotid Stenosis" and "S2e Guideline for Acute Therapy of Ischemic Stroke"). The answers were classified as "correct", "inaccurate", or "incorrect" by two neurovascular experts in consensus. Additionally, retrieval performance of five retrieval strategies was analyzed on a synthetic dataset of 384 questions.

Results: Claude Sonnet 3.5 achieved the highest answer correctness (70.6% correct, 10.6% wrong), followed by Llama 3.1 (64.7%, 15.3% wrong), GPT-4o-mini with RAG (57.6%, 15.3% wrong), and Mixtral (56.6%, 17.6% wrong). GPT-4o-mini without RAG performed significantly worse (20.0%, 32.9% wrong). Retrieval errors were the primary cause of incorrect answers (80%). For retrieval, BM25 achieved the highest accuracy (82.0%), outperforming vector-based methods like "BAAI/bge-m3" (78.4%).

Conclusion: RAG significantly improves LLM accuracy for medical guideline question answering compared to the inherent knowledge of pretrained LLMs alone while still showing significant error rates. Improved accuracy and confidence metrics are needed for safer implementation in clinical routine. Additionally, our results demonstrate the strong performance of general LLMs in medical question answering for non-English languages, such as German, even without specific training.

Purpose: Although flow diverters (FDs) revolutionized the treatment of unruptured intracranial aneurysms (UIAs), clinically silent microinfarcts (SMs) are consistently detected in postprocedural diffusion-weighted imaging (DWI), and their clinical consequences remain unknown. This study examined the incidence of, and factors associated with, SM development after the use of two surface-modified FDs: the p48MW-HPC and p64MW-HPC.

Methods: A non-concurrent cohort study was conducted in 404 consecutive patients with UIAs treated with these two FDs between January 2020 and July 2023. The primary endpoint was the number of asymptomatic DWI lesions within 72 h after the procedure.

Results: SMs were detected in 58.2% of the cases, and ≥ 20 lesions were identified in 7.2% of cases. Multivariate analysis revealed that the use of two specific types of microcatheters-Prowler Select Plus (P < 0.001; IRR, 3.2) and RapidTransit (P = 0.001; IRR, 2.2)-and difficulty in microcatheter navigation to the target vessel (P = 0.039; IRR, 1.8) were significantly associated with greater numbers of lesions. After exclusion of outlier cases, these devices remained associated with a significant 2‑ to 3‑fold increase in SMs. Diabetes was associated with fewer SMs (P = 0.024; IRR, 0.5), although this association disappeared after exclusion of outlier cases.

Conclusion: The postprocedural incidence of SMs associated with hydrophilic polymer-coated FDs was similar to that previously reported. Interestingly, the main factor contributing to this phenomenon was the use of certain microcatheters. Careful device selection is crucial, and further research is needed to improve the safety of neurovascular interventions.