AJNR. American journal of neuroradiology最新文献

{"title":"Intraplaque Hemorrhage Volume and Ischemic Stroke Risk.","authors":"Shahriar Faghani, Mana Moassefi, Erik Albach, Ajay A Madhavan, Ian T Mark, Girish Bathla, Darya P Shlapak, Carrie M Carr, Bradley J Erickson, John C Benson","doi":"10.3174/ajnr.A8889","DOIUrl":"10.3174/ajnr.A8889","url":null,"abstract":"<p><strong>Background and purpose: </strong>Intraplaque hemorrhage (IPH) in carotid atherosclerotic plaques is the best-established biomarker of plaque vulnerability. However, the relationship between IPH volume and ischemic neurologic symptoms remains scarcely studied. This study explored the association between carotid IPH volume and ischemic event severity.</p><p><strong>Materials and methods: </strong>A retrospective analysis was conducted on consecutive patients with suspected carotid atherosclerosis, evaluated from December 2015 to January 2021. Patients underwent carotid plaque MRI using T1-weighted imaging with fat suppression for IPH detection. Included patients had documented neurological symptoms, classified as amaurosis fugax (AF), TIA, and/or stroke. MRI scans were reviewed for presence and volume of IPH, with semi-automated software used for volumetric segmentation. Statistical analyses, including Mann-Whitney U tests and Receiver Operating Characteristic (ROC) curves, were performed to evaluate IPH volume thresholds and their association with symptom severity.</p><p><strong>Results: </strong>The study included 358 patients, of whom 120 had IPH-positive carotid plaques. A higher incidence of ischemic events was noted on the left side, with 28 strokes, 6 AF, and 12 TIAs observed in left-sided events, and 19 strokes, 1 AF, and 3 TIAs in right-sided events. No significant differences in IPH volumes were found across symptom categories or event laterality. ROC analysis identified IPH volume thresholds with AUC values of 0.579 (0.396, 0.748) for left-sided events and 0.618 (0.333, 0.910) for right-sided events, indicating limited discriminatory power for predicting ischemic event severity.</p><p><strong>Conclusions: </strong>While carotid IPH volume is detectable across various neurological symptom categories, our findings indicate that IPH volume alone does not significantly correlate with ischemic event severity. Threshold IPH volumes showed low diagnostic accuracy, suggesting that other plaque characteristics and systemic factors may be more relevant in determining ischemic stroke risk.</p><p><strong>Abbreviations: </strong>IPH＝Intraplaque hemorrhage; AF＝Amaurosis fugax; ROC＝Receiver Operating Characteristic.</p>","PeriodicalId":93863,"journal":{"name":"AJNR. American journal of neuroradiology","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144337253","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Radiologist, Trainee, and Logistical Factors Impacting the Timeliness of CTA Head and Neck Reporting in Stroke Code Activations.","authors":"Omar A Zaree, Jeffers K Nguyen, Irene Dixe de Oliveira Santo, Ahmed E Kertam, Saeed Rahmani, Jason Johnson, Long H Tu","doi":"10.3174/ajnr.A8660","DOIUrl":"10.3174/ajnr.A8660","url":null,"abstract":"<p><strong>Background and purpose: </strong>Timely reporting of CTA examinations impacts management of acute vascular pathology such as large vessel occlusions, arterial dissection, and ruptured aneurysms, as well as a variety of acute nonvascular pathologies. In this study, we examine potential modifiable factors impacting the timeliness of CTA reporting performed in stroke code activations.</p><p><strong>Materials and methods: </strong>This is an observational study of stroke code CTA head and neck examinations performed at a single health system (3 emergency departments, 1550 inpatient beds) during 4 years (January 1, 2019, to December 31, 2023). Patient age, patient sex, care setting, time of year, shift type, trainee/attending radiologist characteristics, report factors, and number of CTAs performed within the preceding hour were considered potential factors impacting the turnaround time (TAT) of stroke code CTAs. Descriptive statistics, univariate regression, and multivariate regression were used to estimate the impact on reporting TAT.</p><p><strong>Results: </strong>We performed 8422 stroke code CTA examinations. Median TAT was 29 minutes (interquartile range [IQR] 18-48). Median TAT by individual attending radiologists varied from 15 to 40 minutes (median of medians, 29 minutes [IQR 26-34.5]). Univariate regression analyses found younger patient age, emergency department setting, time later in the academic year, nonbusiness hours, specific individual radiologists/trainees, solo reporting by attending radiologists, use of preliminary reports, and fewer stroke codes within the preceding hour to all be associated with shorter TATs (all <i>P</i> < .05). After adjusting for patient-, logistical-, and radiologist-level factors in a multivariate regression model, the greatest impact on TAT was seen with variation in individual attending radiologists (adjusted coefficients, -11.9 to +29.4 minutes) and trainees (-40.1 to +95.7 minutes); reporting CTAs without a trainee and release of preliminary reports before final sign were associated with faster TATs (-19.9 and -26.9 minutes, respectively). Each stroke CTA within the preceding hour was associated with only a 2.8-minute increase in TAT. Secondary analyses suggested that previewing of cases during active scanning and use of \"structured\" reports correlate with a favorable impact on TAT among attending radiologists (both <i>P</i> < .05).</p><p><strong>Conclusions: </strong>Radiologist and trainee-level timeliness in stroke CTA reporting varies widely. Interventions aimed at improving workflow efficiency for both trainees and attending radiologists could improve timeliness of reporting.</p>","PeriodicalId":93863,"journal":{"name":"AJNR. American journal of neuroradiology","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143017820","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Frontal Paraventricular Cysts: Refined Definitions and Outcomes.","authors":"Matthew T Whitehead, Amirreza Manteghinejad, César A P F Alves, Onur Simsek, Nahla Khalek, Erin S Schwartz","doi":"10.3174/ajnr.A8653","DOIUrl":"10.3174/ajnr.A8653","url":null,"abstract":"<p><strong>Background and purpose: </strong>Frontal paraventricular cystic changes have a varied etiology that includes connatal cysts, subependymal pseudocysts, necrosis, and enlarged perivascular spaces. These may be difficult to distinguish by neuroimaging and have a variety of associated prognoses. We aim to refine the neuroimaging definition of frontal horn cysts and correlate it with adverse clinical conditions.</p><p><strong>Materials and methods: </strong>In this cross-sectional study, the pre- and postnatal neuroimaging database at a quaternary referral children's hospital was searched for all reports containing \"frontal horn cysts,\" \"periventricular cysts,\" or \"connatal cysts\" after internal review board exemption. Frontal paraventricular abnormalities were categorized as either cysts, necroses, enlarged perivascular spaces, caudothalamic groove subependymal pseudocysts, frontal horn asymmetries, intraventricular septations, or ependymal vessels based on location and appearance. Cyst number, size, location, morphology, and signal/attenuation/echotexture were documented, as were additional brain abnormalities. Clinical outcomes were recorded when available. Fisher exact and χ² tests were used to evaluate categoric data associations and Kruskall-Wallis tests were employed to compare the medians among groups.</p><p><strong>Results: </strong>Two hundred five brain imaging examinations (148 MRI; 55 ultrasound [US]; 2 CT) from 110 distinct subjects (5 fetal: median 29.3, mean 27.5, and range 22.4 to 32.8 gestational weeks; 105 postnatal: mean 2.5 years, median 15 days, range 0 days to 19 years) were included. Seventy-one examinations (35%) were initially diagnosed as connatal cysts but, instead, represented necrosis (<i>n</i> = 23), enlarged perivascular spaces (<i>n</i> = 20), caudothalamic groove germinolytic cysts (<i>n</i> = 11), septations/adhesions (<i>n</i> = 10), ventricular asymmetries (<i>n</i> = 6), and a blood vessel (<i>n</i> = 1). These entities differed in size, shape, location, and orientation (<i>P</i> < .001). Congenital heart disease (<i>P</i> < .04) and gastrointestinal (<i>P</i> < .04) disorders were more common in subjects with frontal cysts and necrosis than in subjects with enlarged perivascular spaces; however, the frontal cyst and necrosis groups showed no differences in outcome (<i>P</i> > .05).</p><p><strong>Conclusions: </strong>Frontal paraventricular cystic changes represent a common interpretive dilemma. Enlarged perivascular spaces should be distinguished from other frontal cystic changes, which portend a more guarded prognosis, whether necrotic or otherwise.</p>","PeriodicalId":93863,"journal":{"name":"AJNR. American journal of neuroradiology","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142967372","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Optimization of Photon-Counting CT Myelography for the Detection of CSF-Venous Fistulas Using Convolutional Neural Network Denoising: A Comparative Analysis of Reconstruction Techniques.","authors":"Ajay A Madhavan, Zhongxing Zhou, Paul J Farnsworth, Jamison Thorne, Timothy J Amrhein, Peter G Kranz, Waleed Brinjikji, Jeremy K Cutsforth-Gregory, Michelle L Kodet, Nikkole M Weber, Grace Thompson, Felix E Diehn, Lifeng Yu","doi":"10.3174/ajnr.A8695","DOIUrl":"10.3174/ajnr.A8695","url":null,"abstract":"<p><strong>Background and purpose: </strong>Photon-counting detector CT myelography (PCD-CTM) is a recently described technique used for detecting spinal CSF leaks, including CSF-venous fistulas. Various image reconstruction techniques, including smoother-versus-sharper kernels and virtual monoenergetic images, are available with photon-counting CT. Moreover, denoising algorithms have shown promise in improving sharp kernel images. No prior studies have compared image quality of these different reconstructions on photon-counting CT myelography. Here, we sought to compare several image reconstructions using various parameters important for the detection of CSF-venous fistulas.</p><p><strong>Materials and methods: </strong>We performed a retrospective review of all consecutive decubitus PCD-CTM between February 1, 2022, and August 1, 2024, at 1 institution. We included patients whose studies had the following reconstructions: Br48-40 keV virtual monoenergetic reconstruction, Br56 low-energy threshold (T3D), Qr89-T3D denoised with quantum iterative reconstruction, and Qr89-T3D denoised with a convolutional neural network algorithm. We excluded patients who had extradural CSF on preprocedural imaging or a technically unsatisfactory myelogram-. All 4 reconstructions were independently reviewed by 2 neuroradiologists. Each reviewer rated spatial resolution, noise, the presence of artifacts, image quality, and diagnostic confidence (whether positive or negative) on a 1-5 scale. These metrics were compared using the Friedman test. Additionally, noise and contrast were quantitatively assessed by a third reviewer and compared.</p><p><strong>Results: </strong>The Qr89 reconstructions demonstrated higher spatial resolution than their Br56 or Br48-40keV counterparts. Qr89 with convolutional neural network denoising had less noise, better image quality, and improved diagnostic confidence compared with Qr89 with quantum iterative reconstruction denoising. The Br48-40keV reconstruction had the highest contrast-to-noise ratio quantitatively.</p><p><strong>Conclusions: </strong>In our study, the sharpest quantitative kernel (Qr89-T3D) with convolutional neural network denoising demonstrated the best performance regarding spatial resolution, noise level, image quality, and diagnostic confidence for detecting or excluding the presence of a CSF-venous fistula.</p>","PeriodicalId":93863,"journal":{"name":"AJNR. American journal of neuroradiology","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143392673","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Systematic Review of the predictive value of negative brain or low probability brain MRIs in patients with CSF venous fistulas.","authors":"Angelique Sao-Mai S Tay, Marcel M Maya, Peter G Kranz, Ajay A Madhavan, Wouter I Schievink","doi":"10.3174/ajnr.A8884","DOIUrl":"https://doi.org/10.3174/ajnr.A8884","url":null,"abstract":"<p><strong>Background: </strong>Since the discovery of the cerebrospinal fluid venous fistula, its diagnosis has become more frequent, especially in patients with brain MRIs positive for spontaneous intracranial hypotension (SIH). However, there is a need to understand the likelihood of diagnosis of a cerebrospinal fluid venous fistula in a patient with negative brain imaging.</p><p><strong>Purpose: </strong>Our aim was to investigate the frequency of cerebrospinal fluid venous fistula in patients suspected of SIH who have negative neuroaxis MRIs.</p><p><strong>Data sources: </strong>All studies reporting on the incidence of cerebrospinal fluid venous fistula in patients with negative neuroaxis MRIs or low probability scores according to the Bern and Mayo score were searched on PubMed, EMBASE, Scopus, Web of Science and Cochrane.</p><p><strong>Study selection: </strong>Nine studies comprising of 898 patients suspected of SIH with 80 cerebrospinal fluid venous fistulas were included.</p><p><strong>Data analysis: </strong>Data were collected on patient demographics, number of patients found to have negative neuroaxis MRIs or low probability scores according to the Bern or Mayo scoring systems, type of imaging used, and number of patients diagnosed with cerebrospinal fluid venous fistula. Analysis was performed using the standard method for evaluating the negative predictive value of a diagnostic test.</p><p><strong>Data synthesis: </strong>There were 27 (10.7%) patients with a cerebrospinal fluid venous fistula of 252 patients found to have negative brain MRIs, 15 (18.3%) of 82 patients found to have low probability on the Bern score, and 38 (34.8%) of 109 patients found to have low probability on the Mayo score. The negative predictive value of a negative brain MRI was 0.89 (95%CI, 0.86-0.92), 0.81 (95% CI, 0.77-0.87) for the Bern score, and 0.65 (95% CI, 0.58-0.72) for the Mayo score.</p><p><strong>Limitations: </strong>Our review was limited by heterogeneity of the reference standard and few studies in each subcategory.</p><p><strong>Conclusions: </strong>This review demonstrated that a negative brain MRI is effective in predicting that a patient will not have a CVF, with a high NPV of 89%. However, a patient with a strong clinical suspicion for CSF leak should be considered for more invasive imaging.</p><p><strong>Abbreviations: </strong>bMRI-- negative brain magnetic resonance imaging, CTM - computed tomography myelogram, CVF - cerebrospinal fluid venous fistula, dCTM-BT - lateral decubitus dynamic CTM protocol using real-time bolus-tracking, DSM - digital subtraction myelogram, NPV - negative predictive value, PC-CTM - photon-counting detector CT myelography.</p>","PeriodicalId":93863,"journal":{"name":"AJNR. American journal of neuroradiology","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144328081","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Can CTA-based Machine Learning Identify Patients for Whom Successful Endovascular Stroke Therapy is Insufficient?","authors":"Jerome A Jeevarajan, Yingjun Dong, Anjan Ballekere, Sergio Salazar Marioni, Arash Niktabe, Rania Abdelkhaleq, Sunil A Sheth, Luca Giancardo","doi":"10.3174/ajnr.A8885","DOIUrl":"https://doi.org/10.3174/ajnr.A8885","url":null,"abstract":"<p><strong>Background and purpose: </strong>Despite advances in endovascular stroke therapy (EST) devices and techniques, many patients are left with substantial disability, even if the final infarct volumes (FIVs) remain small. Here, we evaluate the performance of a machine learning (ML) approach using pre-treatment CT angiography (CTA) to identify this cohort of patients that may benefit from additional interventions.</p><p><strong>Materials and methods: </strong>We identified consecutive large vessel occlusion (LVO) acute ischemic stroke (AIS) subjects who underwent EST with successful reperfusion in a multicenter prospective registry cohort. We included only subjects with FIV<30mL and recorded 90-day outcome (modified Rankin scale, mRS). A deep learning model was pre-trained and then fine-tuned to predict 90-day mRS 0-2 using pre-treatment CTA images (DSN-CTA model). The primary outcome was the predictive performance of the DSNCTA model compared to a logistic regression model with clinical variables, measured by the area under the receiver operating characteristic curve (AUROC).</p><p><strong>Results: </strong>The DSN-CTA model was pre-trained on 1,542 subjects and then fine-tuned and cross-validated with 48 subjects, all of whom underwent EST with TICI 2b-3 reperfusion. Of this cohort, 56.2% of subjects had 90-day mRS 3-6 despite successful EST and FIV<30mL. The DSN-CTA model showed significantly better performance than a model with clinical variables alone when predicting good 90-day mRS (AUROC 0.81 vs 0.492, p=0.006).</p><p><strong>Conclusions: </strong>The CTA-based machine learning model was able to more reliably predict unexpected poor functional outcome after successful EST and small FIV for patients with LVO AIS compared to standard clinical variables. ML models may identify <i>a priori</i> patients in whom EST-based LVO reperfusion alone is insufficient to improve clinical outcomes.</p><p><strong>Abbreviations: </strong>AIS＝ acute ischemic stroke; AUROC＝ area under the receiver operating characteristic curve; DSN-CTA＝ DeepSymNet-v3 model; EST＝ endovascular stroke therapy; FIV＝ final infarct volume; LVO＝ large vessel occlusion; ML＝ machine learning.</p>","PeriodicalId":93863,"journal":{"name":"AJNR. American journal of neuroradiology","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144328071","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Application of Convolutional Neural Network Denoising to Improve Cone Beam CT Myelographic Images.","authors":"Ajay A Madhavan, Zhongxing Zhou, Jamison Thorne, Michelle L Kodet, Jeremy K Cutsforth-Gregory, Wouter I Schievink, Ian T Mark, Beth A Schueler, Lifeng Yu","doi":"10.3174/ajnr.A8877","DOIUrl":"https://doi.org/10.3174/ajnr.A8877","url":null,"abstract":"<p><p>Cone beam CT is an imaging modality that provides high-resolution, cross-sectional imaging in the fluoroscopy suite. In neuroradiology, cone beam CT has been used for various applications including temporal bone imaging and during spinal and cerebral angiography. Furthermore, cone beam CT has been shown to improve imaging of spinal CSF leaks during myelography. One drawback of cone beam CT is that images have a relatively high noise level. In this technical report, we describe the first application of a high-resolution convolutional neural network to denoise cone beam CT myelographic images. We show examples of the resulting improvement in image quality for a variety of types of spinal CSF leaks. Further application of this technique is warranted to demonstrate its clinical utility and potential use for other cone beam CT applications.ABBREVIATIONS: CBCT = cone beam CT; CB-CTM = cone beam CT myelography; CTA = CT angiography; CVF = CSF-venous fistula; DSM = digital subtraction myelography; EID = energy integrating detector; FBP = filtered back-projection; SNR = signal-to-noise ratio.</p>","PeriodicalId":93863,"journal":{"name":"AJNR. American journal of neuroradiology","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-06-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144318928","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"7T MR Angiography for Distinguishing Small Intracranial Aneurysms from Variant Anatomy: Protocols and Impact.","authors":"Vishal Patel, Ahmed K Ahmed, Jorge Rios-Zermeno, Xiangzhi Zhou, Shengzhen Tao, Erin M Westerhold, W David Freeman, Rabih G Tawk, Sukhwinder J S Sandhu, Erik H Middlebrooks","doi":"10.3174/ajnr.A8875","DOIUrl":"https://doi.org/10.3174/ajnr.A8875","url":null,"abstract":"<p><strong>Background and purpose: </strong>Unruptured intracranial aneurysms are increasingly detected on noninvasive imaging, but false positives from limited resolution can lead to unnecessary anxiety, follow-up, and invasive procedures. We investigated multiple 7T MRA sequences for their ability to reduce aneurysm overdiagnosis by differentiating them from variant anatomy. We also evaluated which characteristics of suspected aneurysms were associated with a greater likelihood of diagnostic reversal by 7T MRA and estimated the resulting impact on imaging utilization and cost.</p><p><strong>Materials and methods: </strong>In this retrospective study, 41 suspected aneurysms in 34 patients who underwent 7T MRA over a 22-month period were evaluated using three sequences: conventional TOF, a compressed sensing version of TOF with improved spatial resolution, and contrast-enhanced MRA. Patient demographics, aneurysm size, and prior imaging modality were recorded. Two neuroradiologists assessed each lesion for reclassification as an anatomical variant based on the 7T appearance. Logistic regression was used to identify any significant relationships between the 7T sequence type or aneurysm characteristics and the likelihood of downgrade.</p><p><strong>Results: </strong>Overall, 7T MRA permitted diagnostic downgrade in 46% of suspected aneurysms. Downgrade rates were 30% for conventional TOF, 41% for compressed sensing TOF, and 39% for contrast-enhanced MRA, with no single sequence proving statistically superior. Lesions detected on 1.5T MRA were significantly more likely to be downgraded compared to those found with 3T MRA (53% vs 38%, p < 0.05, OR 2.53). Additionally, aneurysm size was significantly inversely related to downgrade likelihood, with all lesions <1 mm and 63% of lesions 1-2 mm being reclassified, whereas none of the lesions >3 mm were downgraded (p < 0.001, OR 0.30 per mm increase in size, 95%CI 0.15-0.58). Based on these findings, we estimate that 7T MRA can reduce unnecessary surveillance by up to 2.08 scans per patient-resulting in cost savings of up to $1388 per patient depending on the surveillance modality employed and assuming the federal reimbursement rate.</p><p><strong>Conclusions: </strong>7T MRA frequently reclassifies small suspected aneurysms as anatomic variants, especially in cases identified by lower field strength imaging and in smaller lesions. The associated potential for reducing unnecessary follow-up imaging has important cost-saving implications.</p><p><strong>Abbreviations: </strong>CMS = Centers for Medicare and Medicaid Services; CS = compressed sensing; CE = contrast enhanced; UIA = unruptured intracranial aneurysm.</p>","PeriodicalId":93863,"journal":{"name":"AJNR. American journal of neuroradiology","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-06-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144310932","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Predictive Value of CT Perfusion in Spatially and Volumetrically Identifying Ischemic Penumbra Against Final Infarct Size in Anterior Circulation Stroke With and Without Successful Reperfusion.","authors":"Quirin D Strotzer, Rehab N Khalid, Sara De Giorgi, Aman B Patel, Michael H Lev, Rajiv Gupta, Robert W Regenhardt","doi":"10.3174/ajnr.A8876","DOIUrl":"https://doi.org/10.3174/ajnr.A8876","url":null,"abstract":"<p><strong>Background and purpose: </strong>CT perfusion is important for acute ischemic stroke imaging and treatment, but defining the ischemic penumbra and infarct core remains debated. This study examined the relationship between initial hypoperfused tissue and final infarct in acute anterior circulation occlusion patients, stratified by reperfusion status, to assess CT perfusion's predictive spatial and volumetric accuracy and re-evaluate its role in prognostication, nuanced clinical decision making, and novel pathophysiology research in the new large core era.</p><p><strong>Materials and methods: </strong>This retrospective single-center study included patients diagnosed with anterior circulation acute ischemic stroke who underwent CT perfusion between 05/2021-02/2024. Initial imaging and follow-up non-contrast head CT scans were automatically co-registered. CT perfusion-derived ischemic core and penumbra segmentations were compared to manually segmented final infarcts on follow-up imaging. Groups with (endovascular thrombectomy [EVT] with modified Treatment in Cerebral Infarction score ≥ 2b) and without successful reperfusion (no or unsuccessful EVT) were compared using the Dice similarity coefficient for spatial alignment and Spearman correlation for volumetric agreement.</p><p><strong>Results: </strong>One hundred twenty-one patients were included (51% males, median age: 70 years, median NIHSS score: 14). Successful reperfusion was achieved in 37 patients, while 84 received no or unsuccessful EVT. Significant differences in regions of hypoperfusion were observed between the non-reperfusion and reperfusion groups concerning spatial alignment (median Dice similarity coefficient:0.52 vs. 0.29, P < 0.001) and volumetric agreement (Spearman correlation between predicted and actual infarct volumes: 0.72 vs. 0.07, P < 0.001).</p><p><strong>Conclusions: </strong>CT perfusion offers valuable insights into salvageable tissue in acute ischemic stroke, guiding treatments, predicting infarct volumes, aiding in future stroke pathophysiology research in the large core era, and potentially improving prognostication.</p><p><strong>Abbreviations: </strong>EVT = Endovascular Thrombectomy; IQR = Interquartile Range; mTICI = modified Thrombolysis In Cerebral Infarction; rCBF = relative Cerebral Blood Flow.</p>","PeriodicalId":93863,"journal":{"name":"AJNR. American journal of neuroradiology","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-06-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144310933","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Machine Learning-Based Prediction of Delayed Neurological Sequelae in Carbon Monoxide Poisoning Using Automatically Extracted MR Imaging Features.","authors":"Grace Yoojin Lee, Chang Hwan Sohn, Dongwon Kim, Sang-Beom Jeon, Jihye Yun, Sungwon Ham, Yoojin Nam, Jieun Yum, Won Young Kim, Namkug Kim","doi":"10.3174/ajnr.A8870","DOIUrl":"https://doi.org/10.3174/ajnr.A8870","url":null,"abstract":"<p><strong>Background and purpose: </strong>Delayed neurological sequelae are among the most serious complications of carbon monoxide poisoning. However, no reliable tools are available for evaluating its potential risk. We aimed to assess whether machine learning models using imaging features that were automatically extracted from brain MRI can predict the potential delayed neurological sequelae risk in patients with acute carbon monoxide poisoning.</p><p><strong>Materials and methods: </strong>This single-center, retrospective, observational study analyzed a prospectively collected registry of acute carbon monoxide poisoning patients who visited our emergency department from April 2011 to December 2015. Overall, 1618 radiomics and 4 lesion-segmentation features from DWI b1000 and ADC images, as well as 62 clinical variables were extracted from each patient. The entire dataset was divided into five subsets, with one serving as the hold-out test set and the remaining four used for training and tuning. Four machine learning models, linear regression, support vector machine, random forest, and extreme gradient boosting, as well as an ensemble model, were trained and evaluated using 20 different data configurations. The primary evaluation metric was the mean and 95% CI of the area under the receiver operating characteristic curve. Shapley additive explanations were calculated and visualized to enhance model interpretability.</p><p><strong>Results: </strong>Of the 373 patients, delayed neurological sequelae occurred in 99 (26.5%) patients (mean age 43.0 ± 15.2; 62.0% male). The means [95% CIs] of the area under the receiver operating characteristic curve, accuracy, sensitivity, and specificity of the best performing machine learning model for predicting the development of delayed neurological sequelae were 0.88 [0.86-0.9], 0.82 [0.8-0.83], 0.81 [0.79-0.83], and 0.82 [0.8-0.84], respectively. Among imaging features, the presence, size, and number of acute brain lesions on DWI b1000 and ADC images more accurately predicted DNS risk than advanced radiomics features based on shape, texture and wavelet transformation.</p><p><strong>Conclusions: </strong>Machine learning models developed using automatically extracted brain MRI features with clinical features can distinguish patients at delayed neurological sequelae risk. The models enable effective prediction of delayed neurological sequelae in patients with acute carbon monoxide poisoning, facilitating timely treatment planning for prevention.</p><p><strong>Abbreviations: </strong>ABL = Acute brain lesion; AUROC = area under the receiver operating characteristic curve; CO = carbon monoxide; DNS = delayed neurological sequelae; LR = logistic regression; ML = machine learning; RF = random forest; SVM = support vector machine; XGBoost = extreme gradient boosting.</p>","PeriodicalId":93863,"journal":{"name":"AJNR. American journal of neuroradiology","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144287442","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}