Association of Carotid Artery Disease with Collateralization and Infarct Growth in Patients with Acute Middle Cerebral Artery Occlusion.

BACKGROUND AND PURPOSE: Collaterals are important in large vessel occlusions (LVO), but the role of carotid artery disease (CAD) in this context remains unclear. This study aimed to investigate the impact of CAD on intracranial collateralization and infarct growth after thrombectomy in LVO. MATERIALS AND METHODS: All patients who underwent thrombectomy due to M1 segment occlusion from 01/2015 to 12/2021 were retrospectively included. Internal carotid artery stenosis according to NASCET was assessed on the affected and nonaffected sides. Collaterals were assessed according to the Tan score. Infarct growth was quantified by comparing ASPECTS on follow-up imaging with baseline ASPECTS. RESULTS: In total, 709 patients were included, 118 (16.6%) of whom presented with CAD (defined as severe stenosis ≥70% or occlusion ipsilaterally), with 42 cases (5.9%) being contralateral. Good collateralization (Tan 3) was present in 56.5% of the patients with ipsilateral CAD and 69.1% of the patients with contralateral CAD. The ipsilateral stenosis grade was an independent predictor of good collateral supply (adjusted OR: 1.01; NASCET point, 95% CI: 1.00-1.01; P = .009), whereas the contralateral stenosis grade was not (P = .34). Patients with ipsilateral stenosis of ≥70% showed less infarct growth (median ASPECTS decay: 1; IQR: 0-2) compared with patients with 0%-69% stenosis (median: 2; IQR: 1-3) (P = .005). However, baseline ASPECTS was significantly lower in patients with stenosis of 70%-100% (P < .001). The results of a multivariate analysis revealed that increasing ipsilateral stenosis grade (adjusted OR: 1.0; 95% CI: 0.99-1.00; P = .004) and good collateralization (adjusted OR: 0.5; 95% CI: 0.4-0.62; P < .001) were associated with less infarct growth. CONCLUSIONS: CAD of the ipsilateral ICA is an independent predictor of good collateral supply. Patients with CAD tend to have larger baseline infarct size but less infarct growth.

Diffusion tensor imaging in anisotropic tissues: application of reduced gradient vector schemes in peripheral nerves.

BACKGROUND: In contrast to the brain, fibers within peripheral nerves have distinct monodirectional structure questioning the necessity of complex multidirectional gradient vector schemes for DTI. This proof-of-concept study investigated the diagnostic utility of reduced gradient vector schemes in peripheral nerve DTI. METHODS: Three-Tesla magnetic resonance neurography of the tibial nerve using 20-vector DTI (DTI20) was performed in 10 healthy volunteers, 12 patients with type 2 diabetes, and 12 age-matched healthy controls. From the full DTI20 dataset, three reduced datasets including only two or three vectors along the x- and/or y- and z-axes were built to calculate major parameters. The influence of nerve angulation and intraneural connective tissue was assessed. The area under the receiver operating characteristics curve (ROC-AUC) was used for analysis. RESULTS: Simplified datasets achieved excellent diagnostic accuracy equal to DTI20 (ROC-AUC 0.847-0.868, p ≤ 0.005), but compared to DTI20, the reduced models yielded mostly lower absolute values of DTI scalars: median fractional anisotropy (FA) ≤ 0.12; apparent diffusion coefficient (ADC) ≤ 0.25; axial diffusivity ≤ 0.96, radial diffusivity ≤ 0.07). The precision of FA and ADC with the three-vector model was closest to DTI20. Intraneural connective tissue was negatively correlated with FA and ADC (r ≥ -0.49, p < 0.001). Small deviations of nerve angulation had little effect on FA accuracy. CONCLUSIONS: In peripheral nerves, bulk tissue DTI metrics can be approximated with only three predefined gradient vectors along the scanner's main axes, yielding similar diagnostic accuracy as a 20-vector DTI, resulting in substantial scan time reduction. RELEVANCE STATEMENT: DTI bulk tissue parameters of peripheral nerves can be calculated with only three predefined gradient vectors at similar diagnostic performance as a standard DTI but providing a substantial scan time reduction. KEY POINTS: • In peripheral nerves, DTI parameters can be approximated using only three gradient vectors. • The simplified model achieves a similar diagnostic performance as a standard DTI. • The simplified model allows for a significant acceleration of image acquisition. • This can help to introduce multi-b-value DTI techniques into clinical practice.

Advancing noninvasive glioma classification with diffusion radiomics: Exploring the impact of signal intensity normalization.

Background: This study investigates the influence of diffusion-weighted Magnetic Resonance Imaging (DWI-MRI) on radiomic-based prediction of glioma types according to molecular status and assesses the impact of DWI intensity normalization on model generalizability. Methods: Radiomic features, compliant with image biomarker standardization initiative standards, were extracted from preoperative MRI of 549 patients with diffuse glioma, known IDH, and 1p19q-status. Anatomical sequences (T1, T1c, T2, FLAIR) underwent N4-Bias Field Correction (N4) and WhiteStripe normalization (N4/WS). Apparent diffusion coefficient (ADC) maps were normalized using N4 or N4/z-score. Nine machine-learning algorithms were trained for multiclass prediction of glioma types (IDH-mutant 1p/19q codeleted, IDH-mutant 1p/19q non-codeleted, IDH-wild type). Four approaches were compared: Anatomical, anatomical + ADC naive, anatomical + ADC N4, and anatomical + ADC N4/z-score. The University of California San Francisco (UCSF)-glioma dataset (n = 409) was used for external validation. Results: Naïve-Bayes algorithms yielded overall the best performance on the internal test set. Adding ADC radiomics significantly improved AUC from 0.79 to 0.86 (P = .011) for the IDH-wild-type subgroup, but not for the other 2 glioma subgroups (P > .05). In the external UCSF dataset, the addition of ADC radiomics yielded a significantly higher AUC for the IDH-wild-type subgroup (P ≤ .001): 0.80 (N4/WS anatomical alone), 0.81 (anatomical + ADC naive), 0.81 (anatomical + ADC N4), and 0.88 (anatomical + ADC N4/z-score) as well as for the IDH-mutant 1p/19q non-codeleted subgroup (P < .012 each). Conclusions: ADC radiomics can enhance the performance of conventional MRI-based radiomic models, particularly for IDH-wild-type glioma. The benefit of intensity normalization of ADC maps depends on the type and context of the used data.

Prasugrel Single Antiplatelet Therapy versus Aspirin and Clopidogrel Dual Antiplatelet Therapy for Flow Diverter Treatment for Cerebral Aneurysms: A Retrospective Multicenter Study.

BACKGROUND AND PURPOSE: The optimal antiplatelet regimen after flow diverter treatment of cerebral aneurysms is still a matter of debate. A single antiplatelet therapy might be advantageous in determined clinical scenarios. This study evaluated the efficacy and safety of prasugrel single antiplatelet therapy versus aspirin and clopidogrel dual antiplatelet therapy. MATERIALS AND METHODS: We performed a post hoc analysis of 4 retrospective multicenter studies including ruptured and unruptured aneurysms treated with flow diversion using either prasugrel single antiplatelet therapy or dual antiplatelet therapy. Primary end points were the occurrence of any kind of procedure- or device-related thromboembolic complications and complete aneurysm occlusion at the latest radiologic follow-up (mean, 18 months). Dichotomized comparisons of outcomes were performed between single antiplatelet therapy and dual antiplatelet therapy. Additionally, the influence of various patient- and aneurysm-related variables on the occurrence of thromboembolic complications was investigated using multivariable backward logistic regression. RESULTS: A total of 222 patients with 251 aneurysms were included, 90 (40.5%) in the single antiplatelet therapy and 132 (59.5%) in the dual antiplatelet therapy group. The primary outcome-procedure- or device-related thromboembolic complications-occurred in 6 patients (6.6%) of the single antiplatelet therapy and in 12 patients (9.0%) of the dual antiplatelet therapy group (P = .62; OR, 0.712; 95% CI, 0.260-1.930). The primary treatment efficacy end point was reached in 82 patients (80.4%) of the single antiplatelet therapy and in 115 patients (78.2%) of the dual antiplatelet therapy group (P = .752; OR, 1.141; 95% CI, 0.599-2.101). Logistic regression showed that non-surface-modified flow diverters (P = .014) and fusiform aneurysm morphology (P = .004) significantly increased the probability of thromboembolic complications. CONCLUSIONS: Prasugrel single antiplatelet therapy after flow diverter treatment may be as safe and effective as dual antiplatelet therapy and could, therefore, be a valid alternative in selected patients. Further prospective comparative studies are required to validate our findings.

Deep-learning-based reconstruction of undersampled MRI to reduce scan times: a multicentre, retrospective, cohort study.

BACKGROUND: The extended acquisition times required for MRI limit its availability in resource-constrained settings. Consequently, accelerating MRI by undersampling k-space data, which is necessary to reconstruct an image, has been a long-standing but important challenge. We aimed to develop a deep convolutional neural network (dCNN) optimisation method for MRI reconstruction and to reduce scan times and evaluate its effect on image quality and accuracy of oncological imaging biomarkers. METHODS: In this multicentre, retrospective, cohort study, MRI data from patients with glioblastoma treated at Heidelberg University Hospital (775 patients and 775 examinations) and from the phase 2 CORE trial (260 patients, 1083 examinations, and 58 institutions) and the phase 3 CENTRIC trial (505 patients, 3147 examinations, and 139 institutions) were used to develop, train, and test dCNN for reconstructing MRI from highly undersampled single-coil k-space data with various acceleration rates (R=2, 4, 6, 8, 10, and 15). Independent testing was performed with MRIs from the phase 2/3 EORTC-26101 trial (528 patients with glioblastoma, 1974 examinations, and 32 institutions). The similarity between undersampled dCNN-reconstructed and original MRIs was quantified with various image quality metrics, including structural similarity index measure (SSIM) and the accuracy of undersampled dCNN-reconstructed MRI on downstream radiological assessment of imaging biomarkers in oncology (automated artificial intelligence-based quantification of tumour burden and treatment response) was performed in the EORTC-26101 test dataset. The public NYU Langone Health fastMRI brain test dataset (558 patients and 558 examinations) was used to validate the generalisability and robustness of the dCNN for reconstructing MRIs from available multi-coil (parallel imaging) k-space data. FINDINGS: In the EORTC-26101 test dataset, the median SSIM of undersampled dCNN-reconstructed MRI ranged from 0·88 to 0·99 across different acceleration rates, with 0·92 (95% CI 0·92-0·93) for 10-times acceleration (R=10). The 10-times undersampled dCNN-reconstructed MRI yielded excellent agreement with original MRI when assessing volumes of contrast-enhancing tumour (median DICE for spatial agreement of 0·89 [95% CI 0·88 to 0·89]; median volume difference of 0·01 cm3 [95% CI 0·00 to 0·03] equalling 0·21%; p=0·0036 for equivalence) or non-enhancing tumour or oedema (median DICE of 0·94 [95% CI 0·94 to 0·95]; median volume difference of -0·79 cm3 [95% CI -0·87 to -0·72] equalling -1·77%; p=0·023 for equivalence) in the EORTC-26101 test dataset. Automated volumetric tumour response assessment in the EORTC-26101 test dataset yielded an identical median time to progression of 4·27 months (95% CI 4·14 to 4·57) when using 10-times-undersampled dCNN-reconstructed or original MRI (log-rank p=0·80) and agreement in the time to progression in 374 (95·2%) of 393 patients with data. The dCNN generalised well to the fastMRI brain dataset, with significant improvements in the median SSIM when using multi-coil compared with single-coil k-space data (p<0·0001). INTERPRETATION: Deep-learning-based reconstruction of undersampled MRI allows for a substantial reduction of scan times, with a 10-times acceleration demonstrating excellent image quality while preserving the accuracy of derived imaging biomarkers for the assessment of oncological treatment response. Our developments are available as open source software and hold considerable promise for increasing the accessibility to MRI, pending further prospective validation. FUNDING: Deutsche Forschungsgemeinschaft (German Research Foundation) and an Else Kröner Clinician Scientist Endowed Professorship by the Else Kröner Fresenius Foundation.

Deep learning aided preoperative diagnosis of primary central nervous system lymphoma.

The preoperative distinction between glioblastoma (GBM) and primary central nervous system lymphoma (PCNSL) can be difficult, even for experts, but is highly relevant. We aimed to develop an easy-to-use algorithm, based on a convolutional neural network (CNN) to preoperatively discern PCNSL from GBM and systematically compare its performance to experienced neurosurgeons and radiologists. To this end, a CNN-based on DenseNet169 was trained with the magnetic resonance (MR)-imaging data of 68 PCNSL and 69 GBM patients and its performance compared to six trained experts on an external test set of 10 PCNSL and 10 GBM. Our neural network predicted PCNSL with an accuracy of 80% and a negative predictive value (NPV) of 0.8, exceeding the accuracy achieved by clinicians (73%, NPV 0.77). Combining expert rating with automated diagnosis in those cases where experts dissented yielded an accuracy of 95%. Our approach has the potential to significantly augment the preoperative radiological diagnosis of PCNSL.

Hypothalamic subregion alterations in anorexia nervosa and obesity: Association with appetite-regulating hormone levels.

OBJECTIVE: Anorexia nervosa (AN) and obesity are weight-related disorders with imbalances in energy homeostasis that may be due to hormonal dysregulation. Given the importance of the hypothalamus in hormonal regulation, we aimed to identify morphometric alterations to hypothalamic subregions linked to these conditions and their connection to appetite-regulating hormones. METHODS: Structural magnetic resonance imaging (MRI) was obtained from 78 patients with AN, 27 individuals with obesity and 100 normal-weight healthy controls. Leptin, ghrelin, and insulin blood levels were measured in a subsample of each group. An automated segmentation method was used to segment the hypothalamus and its subregions. Volumes of the hypothalamus and its subregions were compared between groups, and correlational analysis was employed to assess the relationship between morphometric measurements and appetite-regulating hormone levels. RESULTS: While accounting for total brain volume, patients with AN displayed a smaller volume in the inferior-tubular subregion (ITS). Conversely, obesity was associated with a larger volume in the anterior-superior, ITS, posterior subregions (PS), and entire hypothalamus. There were no significant volumetric differences between AN subtypes. Leptin correlated positively with PS volume, whereas ghrelin correlated negatively with the whole hypothalamus volume in the entire cohort. However, appetite-regulating hormone levels did not mediate the effects of body mass index on volumetric measures. CONCLUSION: Our results indicate the importance of regional structural hypothalamic alterations in AN and obesity, extending beyond global changes to brain volume. Furthermore, these alterations may be linked to changes in hormonal appetite regulation. However, given the small sample size in our correlation analysis, further analyses in a larger sample size are warranted. PUBLIC SIGNIFICANCE: Using an automated segmentation method to investigate morphometric alterations of hypothalamic subregions in AN and obesity, this study provides valuable insights into the complex interplay between hypothalamic alterations, hormonal appetite regulation, and body weight, highlighting the need for further research to uncover underlying mechanisms.

Magnetic resonance neurography in spinal cord injury: Imaging findings and clinical significance.

BACKGROUND AND PURPOSE: It is unknown whether changes to the peripheral nervous system following spinal cord injury (SCI) are relevant for functional recovery or the development of neuropathic pain below the level of injury. Magnetic resonance neurography (MRN) at 3 T allows detection and localization of structural and functional nerve damage. This study aimed to combine MRN and clinical assessments in individuals with chronic SCI and nondisabled controls. METHODS: Twenty participants with chronic SCI and 20 controls matched for gender, age, and body mass index underwent MRN of the L5 dorsal root ganglia (DRG) and the sciatic nerve. DRG volume, sciatic nerve mean cross-sectional area (CSA), fascicular lesion load, and fractional anisotropy (FA), a marker for functional nerve integrity, were calculated. Results were correlated with clinical assessments and nerve conduction studies. RESULTS: Sciatic nerve CSA and lesion load were higher (21.29 ± 5.82 mm2 vs. 14.08 ± 4.62 mm2 , p < 0.001; and 8.70 ± 7.47% vs. 3.60 ± 2.45%, p < 0.001) in individuals with SCI compared to controls, whereas FA was lower (0.55 ± 0.11 vs. 0.63 ± 0.08, p = 0.022). DRG volumes were larger in individuals with SCI who suffered from neuropathic pain compared to those without neuropathic pain (223.7 ± 53.08 mm3 vs. 159.7 ± 55.66 mm3 , p = 0.043). Sciatic MRN parameters correlated with electrophysiological results but did not correlate with the extent of myelopathy or clinical severity of SCI. CONCLUSIONS: Individuals with chronic SCI are subject to a decline of structural peripheral nerve integrity that may occur independently from the clinical severity of SCI. Larger volumes of DRG in SCI with neuropathic pain support existing evidence from animal studies on SCI-related neuropathic pain.

Phase angle of bioelectrical impedance analysis as an indicator for diabetic polyneuropathy in type 2 diabetes mellitus.

CONTEXT: Due to the heterogenous clinical symptoms and deficits, the diagnosis of diabetic polyneuropathy (DPN) is still difficult in clinical routine leading to increased morbidity and mortality. OBJECTIVE: We studied the correlation of phase angle (PhA) of bioelectrical impedance analysis (BIA) with clinical, laboratory and physical markers of DPN to evaluate PhA as possible diagnostic method for DPN. MATERIALS AND METHODS: In this cross-sectional observational study as part of the Heidelberg Study on Diabetes and Complications we examined 104 healthy individuals and 205 patients with type 2 diabetes mellitus (T2D), amongst which 63 had DPN. The PhA was calculated from multi-frequency BIA. Nerve conduction studies (NCS), quantitative sensory testing (QST) and diffusion-weighted magnetic resonance neurography (MRN) to determine fractional anisotropy (FA) reflecting peripheral nerve integrity were performed. RESULTS: T2D patients with DPN had lower PhA values (5.71 ± 0.10) compared to T2D patients without DPN (6.07 ± 0.08, p = 0.007, + 6.1%) and healthy controls (6.18 ± 0.08, p < 0.001, + 7.9%). Confounder-adjusted analyses showed correlations of the PhA with conduction velocities and amplitudes of the peroneal (ß=0.28; ß=0.31, p < 0.001) and tibial nerves (ß=0.28; ß=0.32, p < 0.001), Z-scores of QST (thermal detection ß=0.30, p < 0.05) and the FA (ß=0.60, p < 0.001). ROC analysis showed similar performance of PhA in comparison to mentioned diagnostic methods. CONCLUSION: The study shows that PhA is in comparison to other test systems used, at least an equally good and much easier to handle, investigator independent marker for detection of DPN.

Magnetization transfer ratio of the sciatic nerve differs between patients in type 1 and type 2 diabetes.

BACKGROUND: Previous studies on magnetic resonance neurography (MRN) found different patterns of structural nerve damage in type 1 diabetes (T1D) and type 2 diabetes (T2D). Magnetization transfer ratio (MTR) is a quantitative technique to analyze the macromolecular tissue composition. We compared MTR values of the sciatic nerve in patients with T1D, T2D, and healthy controls (HC). METHODS: 3-T MRN of the right sciatic nerve at thigh level was performed in 14 HC, 10 patients with T1D (3 with diabetic neuropathy), and 28 patients with T2D (10 with diabetic neuropathy). Results were subsequently correlated with clinical and electrophysiological data. RESULTS: The sciatic nerve's MTR was lower in patients with T2D (0.211 ± 0.07, mean ± standard deviation) compared to patients with T1D (T1D 0.285 ± 0.03; p = 0.015) and HC (0.269 ± 0.05; p = 0.039). In patients with T1D, sciatic MTR correlated positively with tibial nerve conduction velocity (NCV; r = 0.71; p = 0.021) and negatively with hemoglobin A1c (r = - 0.63; p < 0.050). In patients with T2D, we found negative correlations of sciatic nerve's MTR peroneal NCV (r = - 0.44; p = 0.031) which remained significant after partial correlation analysis controlled for age and body mass index (r = 0.51; p = 0.016). CONCLUSIONS: Lower MTR values of the sciatic nerve in T2D compared to T1D and HC and diametrical correlations of MTR values with NCV in T1D and T2D indicate that there are different macromolecular changes and pathophysiological pathways underlying the development of neuropathic nerve damage in T1D and T2D. TRIAL REGISTRATION: https://classic. CLINICALTRIALS: gov/ct2/show/NCT03022721 . 16 January 2017. RELEVANCE STATEMENT: Magnetization transfer ratio imaging may serve as a non-invasive imaging method to monitor the diseases progress and to encode the pathophysiology of nerve damage in patients with type 1 and type 2 diabetes. KEY POINTS: • Magnetization transfer imaging detects distinct macromolecular nerve lesion patterns in diabetes patients. • Magnetization transfer ratio was lower in type 2 diabetes compared to type 1 diabetes. • Different pathophysiological mechanisms drive nerve damage in type 1 and 2 diabetes.

MRI Gadolinium-Based Contrast Media: Meeting Radiological, Clinical, and Environmental Needs.

Gadolinium-based contrast agents (GBCAs) are routinely used in magnetic resonance imaging (MRI). They are essential for choosing the most appropriate medical or surgical strategy for patients with serious pathologies, particularly in oncologic, inflammatory, and cardiovascular diseases. However, GBCAs have been associated with an increased risk of nephrogenic systemic fibrosis in patients with renal failure, as well as the possibility of deposition in the brain, bones, and other organs, even in patients with normal renal function. Research is underway to reduce the quantity of gadolinium injected, without compromising image quality and diagnosis. The next generation of GBCAs will enable a reduction in the gadolinium dose administered. Gadopiclenol is the first of this new generation of GBCAs, with high relaxivity, thus having the potential to reduce the gadolinium dose while maintaining good in vivo stability due to its macrocyclic structure. High-stability and high-relaxivity GBCAs will be one of the solutions for reducing the dose of gadolinium to be administered in clinical practice, while the development of new technologies, including optimization of MRI acquisitions, new contrast mechanisms, and artificial intelligence may help reduce the need for GBCAs. Future solutions may involve a combination of next-generation GBCAs and image-processing techniques to optimize diagnosis and treatment planning while minimizing exposure to gadolinium. LEVEL OF EVIDENCE: 5 TECHNICAL EFFICACY: Stage 3.

Sine Spin flat detector CT can improve cerebral soft tissue imaging: a retrospective in vivo study.

BACKGROUND: Flat detector computed tomography (FDCT) is frequently applied for periinterventional brain imaging within the angiography suite. Novel technical developments such as the Sine Spin FDCT (S-FDCT) may provide an improved cerebral soft tissue contrast. This study investigates the effect of S-FDCT on the differentiation between gray and white matter compared to conventional FDCT (C-FDCT) and multidetector computed tomography (MDCT). METHODS: A retrospective analysis of a prospectively maintained patient database was performed, including patients who underwent mechanical thrombectomy in our institution and received S-FDCT or C-FDCT as well as MDCT. Differentiation between gray and white matter on the contralateral hemisphere to the ischemic stroke was analyzed quantitatively by contrast-to-noise ratio (CNR) and qualitatively (5-point ordinal scale). RESULTS: In a cohort of 109 patients, MDCT demonstrated the best differentiation between gray and white matter compared to both FDCT techniques (p ≤ 0.001). Comparing both generations of FDCT, S-FDCT provided better visibility of the basal ganglia (p = 0.045) and the supratentorial cortex (p = 0.044) compared to C-FDCT both in quantitative and qualitative analyses. Median CNR were as follows: S-FDCT 2.41 (interquartile range [IQR] 1.66-3.21), C-FDCT 0.96 (0.46-1.70), MDCT 3.43 (2.83-4.17). For basal ganglia, median score and IQR were as follows: S-FDCT 2.00 (2.00-3.00), C-FDCT 1.50 (1.00-2.00), MDCT 5.00 (4.00-5.00). CONCLUSIONS: The novel S-FDCT improves the periinterventional imaging quality of cerebral soft tissue compared to C-FDCT. Thus, it may improve the diagnosis of complications within the angiography suite. MDCT provides the best option for x-ray-based imaging of the brain tissue. RELEVANCE STATEMENT: Flat detector computed tomography is a promising technique for cerebral soft tissue imaging, while the novel Sine Spin flat detector computed tomography technique improves imaging quality compared to conventional flat detector computed tomography and thus may facilitate periinterventional diagnosis of gray and white matter. KEY POINTS: • Flat detector computed tomography (FDCT) is frequently applied for periinterventional brain imaging. • The potential of novel Sine Spin FDCT (S-FDCT) is unknown so far. • S-FDCT improves the visibility of cerebral soft tissue compared to conventional FDCT. • Multidetector computed tomography is superior to both FDCT techniques. • S-FDCT may facilitate the evaluation of brain parenchyma within the angiography suite.

Deep Learning-based Identification of Brain MRI Sequences Using a Model Trained on Large Multicentric Study Cohorts.

Purpose To develop a fully automated device- and sequence-independent convolutional neural network (CNN) for reliable and high-throughput labeling of heterogeneous, unstructured MRI data. Materials and Methods Retrospective, multicentric brain MRI data (2179 patients with glioblastoma, 8544 examinations, 63 327 sequences) from 249 hospitals and 29 scanner types were used to develop a network based on ResNet-18 architecture to differentiate nine MRI sequence types, including T1-weighted, postcontrast T1-weighted, T2-weighted, fluid-attenuated inversion recovery, susceptibility-weighted, apparent diffusion coefficient, diffusion-weighted (low and high b value), and gradient-recalled echo T2*-weighted and dynamic susceptibility contrast-related images. The two-dimensional-midsection images from each sequence were allocated to training or validation (approximately 80%) and testing (approximately 20%) using a stratified split to ensure balanced groups across institutions, patients, and MRI sequence types. The prediction accuracy was quantified for each sequence type, and subgroup comparison of model performance was performed using χ2 tests. Results On the test set, the overall accuracy of the CNN (ResNet-18) ensemble model among all sequence types was 97.9% (95% CI: 97.6, 98.1), ranging from 84.2% for susceptibility-weighted images (95% CI: 81.8, 86.6) to 99.8% for T2-weighted images (95% CI: 99.7, 99.9). The ResNet-18 model achieved significantly better accuracy compared with ResNet-50 despite its simpler architecture (97.9% vs 97.1%; P ≤ .001). The accuracy of the ResNet-18 model was not affected by the presence versus absence of tumor on the two-dimensional-midsection images for any sequence type (P > .05). Conclusion The developed CNN (www.github.com/neuroAI-HD/HD-SEQ-ID) reliably differentiates nine types of MRI sequences within multicenter and large-scale population neuroimaging data and may enhance the speed, accuracy, and efficiency of clinical and research neuroradiologic workflows. Keywords: MR-Imaging, Neural Networks, CNS, Brain/Brain Stem, Computer Applications-General (Informatics), Convolutional Neural Network (CNN), Deep Learning Algorithms, Machine Learning Algorithms Supplemental material is available for this article. © RSNA, 2023.

PerSurge (NOA-30) phase II trial of perampanel treatment around surgery in patients with progressive glioblastoma.

BACKGROUND: Glioblastoma is the most frequent and a particularly malignant primary brain tumor with no efficacy-proven standard therapy for recurrence. It has recently been discovered that excitatory synapses of the AMPA-receptor subtype form between non-malignant brain neurons and tumor cells. This neuron-tumor network connectivity contributed to glioma progression and could be efficiently targeted with the EMA/FDA approved antiepileptic AMPA receptor inhibitor perampanel in preclinical studies. The PerSurge trial was designed to test the clinical potential of perampanel to reduce tumor cell network connectivity and tumor growth with an extended window-of-opportunity concept. METHODS: PerSurge is a phase IIa clinical and translational treatment study around surgical resection of progressive or recurrent glioblastoma. In this multicenter, 2-arm parallel-group, double-blind superiority trial, patients are 1:1 randomized to either receive placebo or perampanel (n = 66 in total). It consists of a treatment and observation period of 60 days per patient, starting 30 days before a planned surgical resection, which itself is not part of the study interventions. Only patients with an expected safe waiting interval are included, and a safety MRI is performed. Tumor cell network connectivity from resected tumor tissue on single cell transcriptome level as well as AI-based assessment of tumor growth dynamics in T2/FLAIR MRI scans before resection will be analyzed as the co-primary endpoints. Secondary endpoints will include further imaging parameters such as pre- and postsurgical contrast enhanced MRI scans, postsurgical T2/FLAIR MRI scans, quality of life, cognitive testing, overall and progression-free survival as well as frequency of epileptic seizures. Further translational research will focus on additional biological aspects of neuron-tumor connectivity. DISCUSSION: This trial is set up to assess first indications of clinical efficacy and tolerability of perampanel in recurrent glioblastoma, a repurposed drug which inhibits neuron-glioma synapses and thereby glioblastoma growth in preclinical models. If perampanel proved to be successful in the clinical setting, it would provide the first evidence that interference with neuron-cancer interactions may indeed lead to a benefit for patients, which would lay the foundation for a larger confirmatory trial in the future. TRIAL REGISTRATION: EU-CT number: 2023-503938-52-00 30.11.2023.

Long-term Follow-up After Aneurysm Treatment with the Flow Redirection Endoluminal Device (FRED) Flow Diverter.

INTRODUCTION: This study focuses on long-term outcomes after aneurysm treatment with either the Flow Re-Direction Endoluminal Device (FRED) or the FRED Jr. to investigate the durability of treatment effect and long-term complications. METHODS: This study is based on a retrospective analysis of a prospectively maintained patient data base. Patients treated with either FRED or FRED Jr. between 2013 and 2017 at our institution, and thus a possibility for ≥ 5 years of follow-up, were included. Aneurysm occlusion rates, recurrence rates, modified Rankin scale score shifts to baseline, and delayed complications were assessed. RESULTS: In this study 68 patients with 84 aneurysms had long-term follow-up with a mean duration of 57.3 months and 44 patients harboring 52 aneurysms had a follow-up ≥ 5 years with a mean follow-up period of 69.2 months. Complete occlusion was reached in 77.4% at 2 years and increased to 84.9% when the latest available imaging result was considered. Younger age and the absence of branch involvement were predictors for aneurysm occlusion in linear regression analysis. After the 2­year threshold, there were 3 reported symptomatic non-serious adverse events. Of these, one patient had a minor stroke, one a transitory ischemic attack and one had persistent mass effect symptoms due to a giant aneurysm, none of these resulted in subsequent neurological disability. CONCLUSION: This long-term follow-up study demonstrates that the FRED and FRED Jr. are safe and effective for the treatment of cerebral aneurysms in the long term, with high rates of complete occlusion and low rates of delayed adverse events.

Impact of signal intensity normalization of MRI on the generalizability of radiomic-based prediction of molecular glioma subtypes.

OBJECTIVES: Radiomic features have demonstrated encouraging results for non-invasive detection of molecular biomarkers, but the lack of guidelines for pre-processing MRI-data has led to poor generalizability. Here, we assessed the influence of different MRI-intensity normalization techniques on the performance of radiomics-based models for predicting molecular glioma subtypes. METHODS: Preoperative MRI-data from n = 615 patients with newly diagnosed glioma and known isocitrate dehydrogenase (IDH) and 1p/19q status were pre-processed using four different methods: no normalization (naive), N4 bias field correction (N4), N4 followed by either WhiteStripe (N4/WS), or z-score normalization (N4/z-score). A total of 377 Image-Biomarker-Standardisation-Initiative-compliant radiomic features were extracted from each normalized data, and 9 different machine-learning algorithms were trained for multiclass prediction of molecular glioma subtypes (IDH-mutant 1p/19q codeleted vs. IDH-mutant 1p/19q non-codeleted vs. IDH wild type). External testing was performed in public glioma datasets from UCSF (n = 410) and TCGA (n = 160). RESULTS: Support vector machine yielded the best performance with macro-average AUCs of 0.84 (naive), 0.84 (N4), 0.87 (N4/WS), and 0.87 (N4/z-score) in the internal test set. Both N4/WS and z-score outperformed the other approaches in the external UCSF and TCGA test sets with macro-average AUCs ranging from 0.85 to 0.87, replicating the performance of the internal test set, in contrast to macro-average AUCs ranging from 0.19 to 0.45 for naive and 0.26 to 0.52 for N4 alone. CONCLUSION: Intensity normalization of MRI data is essential for the generalizability of radiomic-based machine-learning models. Specifically, both N4/WS and N4/z-score approaches allow to preserve the high model performance, yielding generalizable performance when applying the developed radiomic-based machine-learning model in an external heterogeneous, multi-institutional setting. CLINICAL RELEVANCE STATEMENT: Intensity normalization such as N4/WS or N4/z-score can be used to develop reliable radiomics-based machine learning models from heterogeneous multicentre MRI datasets and provide non-invasive prediction of glioma subtypes. KEY POINTS: • MRI-intensity normalization increases the stability of radiomics-based models and leads to better generalizability. • Intensity normalization did not appear relevant when the developed model was applied to homogeneous data from the same institution. • Radiomic-based machine learning algorithms are a promising approach for simultaneous classification of IDH and 1p/19q status of glioma.

Cortical atrophy on baseline computed tomography imaging predicts clinical outcome in patients undergoing endovascular treatment for acute ischemic stroke.

OBJECTIVE: Multiple variables beyond the extent of recanalization can impact the clinical outcome after acute ischemic stroke due to large vessel occlusions. Here, we assessed the influence of small vessel disease and cortical atrophy on clinical outcome using native cranial computed tomography (NCCT) in a large single-center cohort. METHODS: A total of 1103 consecutive patients who underwent endovascular treatment (EVT) due to occlusion of the middle cerebral artery territory were included. NCCT data were visually assessed for established markers of age-related white matter changes (ARWMC) and brain atrophy. All images were evaluated separately by two readers to assess the inter-observer variability. Regression and machine learning models were built to determine the predictive relevance of ARWMC and atrophy in the presence of important baseline clinical and imaging metrics. RESULTS: Patients with favorable outcome presented lower values for all measured metrics of pre-existing brain deterioration (p < 0.001). Both ARWMC (p < 0.05) and cortical atrophy (p < 0.001) were independent predictors of clinical outcome at 90 days when controlled for confounders in both regression analyses and led to a minor improvement of prediction accuracy in machine learning models (p < 0.001), with atrophy among the top-5 predictors. CONCLUSION: NCCT-based cortical atrophy and ARWMC scores on NCCT were strong and independent predictors of clinical outcome after EVT. CLINICAL RELEVANCE STATEMENT: Visual assessment of cortical atrophy and age-related white matter changes on CT could improve the prediction of clinical outcome after thrombectomy in machine learning models which may be integrated into existing clinical routines and facilitate patient selection. KEY POINTS: • Cortical atrophy and age-related white matter changes were quantified using CT-based visual scores. • Atrophy and age-related white matter change scores independently predicted clinical outcome after mechanical thrombectomy and improved machine learning-based prediction models. • Both scores could easily be integrated into existing clinical routines and prediction models.

Iatrogenic Air Embolisms During Endovascular Interventions: Impact of Origin and Number of Air Bubbles on Cerebral Infarctions.

PURPOSE: Cerebral infarctions caused by air embolisms (AE) are a feared risk in endovascular procedures; however, the relevance and pathophysiology of these AEs is still largely unclear. The objective of this study was to investigate the impact of the origin (aorta, carotid artery or right atrium) and number of air bubbles on cerebral infarctions in an experimental in vivo model. METHODS: In 20 rats 1200 or 2000 highly calibrated micro air bubbles (MAB) with a size of 85 µm were injected at the aortic valve (group Ao), into the common carotid artery (group CA) or into the right atrium (group RA) using a microcatheter via a transfemoral access, resembling endovascular interventions in humans. Magnetic resonance imaging (MRI) using a 9.4T system was performed 1 h after MAB injection followed by finalization. RESULTS: The number (5.5 vs. 5.5 median) and embolic patterns of infarctions did not significantly differ between groups Ao and CA. The number of infarctions were significantly higher comparing 2000 and 1200 injected MABs (6 vs. 4.5; p < 0.001). The infarctions were significantly larger for group CA (median infarction volume: 0.41 mm3 vs. 0.19 mm3; p < 0.001). In group RA and in the control group no infarctions were detected. Histopathological analyses showed early signs of ischemic stroke. CONCLUSION: Iatrogenic AEs originating at the ascending aorta cause a similar number and pattern of cerebral infarctions compared to those with origin at the carotid artery. These findings underline the relevance and potential risk of AE occurring during endovascular interventions at the aortic valve and ascending aorta.

Association of Small Fiber Function with Microvascular Perfusion of Peripheral Nerves in Patients with Type 2 Diabetes : Study using Quantitative Sensory Testing and Magnetic Resonance Neurography.

INTRODUCTION/AIMS: Diabetic small fiber neuropathy (SFN) is caused by damage to thinly myelinated A­fibers (δ) and unmyelinated C­fibers. This study aimed to assess associations between quantitative sensory testing (QST) and parameters of peripheral nerve perfusion obtained from dynamic contrast enhanced (DCE) magnetic resonance neurography (MRN) in type 2 diabetes patients with and without SFN. METHODS: A total of 18 patients with type 2 diabetes (T2D, 8 with SFN, 10 without SFN) and 10 healthy controls (HC) took part in this cross-sectional single-center study and underwent QST of the right leg and DCE-MRN of the right thigh with subsequent calculation of the sciatic nerve constant of capillary permeability (Ktrans), extravascular extracellular volume fraction (Ve), and plasma volume fraction (Vp). RESULTS: The Ktrans (HC 0.031 min-1 ± 0.009, T2D 0.043 min-1 ± 0.015; p = 0.033) and Ve (HC 1.2% ± 1.5, T2D: 4.1% ± 5.1; p = 0.027) were lower in T2D patients compared to controls. In T2D patients, compound z­scores of thermal and mechanical detection correlated with Ktrans (r = 0.73; p = 0.001, and r = 0.57; p = 0.018, respectively) and Ve (r = 0.67; p = 0.002, and r = 0.69; p = 0.003, respectively). Compound z­scores of thermal pain and Vp (r = -0.57; p = 0.015) correlated negatively. DISCUSSION: The findings suggest that parameters of peripheral nerve microcirculation are related to different symptoms in SFN: A reduced capillary permeability may result in a loss of function related to insufficient nutritional supply, whereas increased capillary permeability may be accompanied by painful symptoms related to a gain of function.