AJNR. American journal of neuroradiology最新文献

{"title":"Catheter-Vessel Interactions During Aspiration Thrombectomy: A Comprehensive Analysis Using a Novel Swine Model.","authors":"Timothy K Shimizu, Ichiro Yuki, Hemdeep Kaur, Zachary W Hsu, Earl Steward, Tracy Nguyen, Shuichi Suzuki","doi":"10.3174/ajnr.A8898","DOIUrl":"https://doi.org/10.3174/ajnr.A8898","url":null,"abstract":"<p><strong>Background and purpose: </strong>Catheter-vessel interactions during aspiration thrombectomy may vary depending on vessel size, potentially affecting reperfusion outcomes. This study aimed to assess how vessel diameter influences procedural dynamics and reperfusion success rates using a swine model with real-time, direct vessel visualization.</p><p><strong>Materials and methods: </strong>An animal model utilizing Yorkshire swine was used to evaluate catheter-vessel interactions during aspiration thrombectomy. The common carotid artery (CCA, 4-6 mm) and superficial cervical artery (SCA, 2-3 mm) represented large and small vessels, respectively. Each vessel was surgically exposed, occluded using clot analogs, and monitored using a high-resolution digital microscope camera. Aspiration thrombectomy was performed starting with remote aspiration, followed by contact aspiration. Vessel behavior, including collapse and traction, was evaluated in relation to vessel size and reperfusion success.</p><p><strong>Results: </strong>A total of 33 thrombectomy procedures were performed on 14 vessels in 4 swine. In the CCA group (mean diameter: 5.14 mm), remote aspiration achieved 85.7% reperfusion success without vessel collapse, leading to 100% final reperfusion. In the SCA group (mean diameter: 2.27 mm), all remote aspiration attempts caused immediate vessel collapse with no successful reperfusion (0%), requiring contact aspiration, which resulted in 71.4% final reperfusion. Vessel collapse was significantly more frequent in SCA vs. CCA (p < 0.001). Two distinct patterns of vessel traction were observed exclusively in the SCA group.</p><p><strong>Conclusions: </strong>Vessel size strongly influences aspiration thrombectomy efficacy. Smaller vessels are more susceptible to vessel collapse and catheter-induced traction during aspiration thrombectomy, reducing the success rate of reperfusion and increasing reliance on true contact aspiration.</p><p><strong>Abbreviations: </strong>CCA＝ common carotid artery; SCA＝ superficial cervical artery; LVO＝ large-vessel occlusion.</p>","PeriodicalId":93863,"journal":{"name":"AJNR. American journal of neuroradiology","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144546527","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":"State of Practice: A Report from the Inaugural SNIS Neurointerventional Oncology Summit.","authors":"Christopher C Young, Kazim H Narsinh, Stephen R Chen, Sameer A Ansari, Steven W Hetts, Frederick F Lang, Max Wintermark, Peter T Kan","doi":"10.3174/ajnr.A8902","DOIUrl":"https://doi.org/10.3174/ajnr.A8902","url":null,"abstract":"<p><strong>Background: </strong>Over the past 25-years, progress in the treatment of central nervous system (CNS) tumors has been limited and outcomes for malignancies such as glioblastoma and diffuse intrinsic pontine glioma remain dismal. There has been great interest in harnessing endovascular neurointerventional techniques and to use the cerebral vasculature as a route for therapeutic delivery in neuro-oncology. Several selective intra-arterial clinical trials are currently underway targeting a range of CNS tumors with different therapeutic agents.</p><p><strong>Methods: </strong>The Society of NeuroInterventional Surgery (SNIS) convened the inaugural Neurointerventional Oncology Summit which was held on April 26-27, 2024, at the University of Texas MD Anderson Cancer Center in Houston, Texas. The group consisted of neuro-interventionists, neurosurgeons, neuro-oncologists, neuroradiologists, basic scientists and industry representatives who are current practitioners in this emerging space. We report the current state of practice and our efforts to develop an organizational platform to treat patients suffering from CNS cancers and to advance research in this emerging field.</p><p><strong>Key message: </strong>Endovascular surgical neuro-oncology is an emerging clinical endeavor which aims to provide novel therapeutic options for CNS cancers. Novel therapeutic agents including yttrium-90, oncolytic viruses and cellular immunotherapy are being tested as intraarterial therapy. Technological advances in imaging modalities, selective methods of blood brain barrier opening, and devices and catheters will aid advancement of the field. Multi-disciplinary cooperation and collaboration will be integral to the success of this endeavor.</p><p><strong>Abbreviations: </strong>BBB = blood brain barrier; CAR = chimeric antigen receptor; DIPG = diffuse intrinsic pontine glioma; ESIA = endovascular selective intra-arterial; ESSIA = endovascular super-selective intra-arterial; FUS = focused ultrasound; GBM = glioblastoma; MSC-D24 = delta 24 oncolytic virus in mesenchymal stem cell; NK = natural killer; Y-90 = yttrium-90.</p>","PeriodicalId":93863,"journal":{"name":"AJNR. American journal of neuroradiology","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144546529","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":"Hydrogel Coils versus Bare Platinum Coils for the Treatment of Ruptured and Unruptured Aneurysms: An Updated Systematic Review and Meta-Analysis of Randomized Controlled Trials.","authors":"Jonathan Cortese, Sherief Ghozy, Armin Zarrintan, Delal Bektas, Omar M Al-Janabi, Onam Verma, Esref Alperen Bayraktar, Waleed Brinjikji, Ramanathan Kadirvel, David F Kallmes","doi":"10.3174/ajnr.A8697","DOIUrl":"https://doi.org/10.3174/ajnr.A8697","url":null,"abstract":"<p><strong>Background and purpose: </strong>Whether hydrogel coils (HGCs) can reduce intracranial aneurysm recurrences when compared with bare platinum coils (BPCs) remains a disputed subject. Thus, we sought to perform a systematic review and meta-analysis to evaluate the efficacy of hydrogel coils in the context of intracranial aneurysm treatment.</p><p><strong>Materials and methods: </strong>Following PRISMA 2020 guidelines, we systematically reviewed PubMed, Scopus, Embase, and Web of Science for randomized controlled trials (RCTs) comparing HGC to BPC. Outcomes of interest were: end of procedure and last follow-up occlusion rates, including complete occlusion and major recurrence, complication rates, morbidity, and mortality. Risk ratios (RRs) and 95% CIs were calculated.</p><p><strong>Results: </strong>The study selection: 5 RCTs, including 2126 patients (HGC <i>n</i> = 1064, BPC <i>n</i> = 1062), were analyzed. HGC showed comparable rates of immediate complete occlusion (RR = 0.89, 95% CI = 0.68-1.16, <i>P</i> = .29) and packing attenuation (MD = 27.17, 95% CI = -16.59-70.93, <i>P</i> = .12) compared with BPC. At an average 18-month follow-up, HGC significantly reduced major recurrence rates (RR = 0.75, 95% CI = 0.60-0.94, <i>P</i> = .03). Complete occlusion rates at the last follow-up were higher for HGC after outlier exclusion (RR = 1.29, 95% CI = 1.18-1.42, <i>P</i> < .001). All the outcomes related to complications, including hemorrhagic and thromboembolic complications, were similar between both groups (<i>P</i> > .1 for all). Finally, HGC resulted in similar rates of mRS 0-2 and mortality compared with BPC (RR = 0.98, 95% CI = 0.95-1.01, <i>P</i> = .15 and RR = 0.72, 95% CI = 0.31-1.65, <i>P</i> = .33, respectively). Only 5 RCTs were included in this meta-analysis, which may limit the generalizability of our findings. The absence of long-term follow-up also limits the assessment of treatment durability.</p><p><strong>Conclusions: </strong>Our meta-analysis of RCTs suggests that the use of HGC in the endovascular treatment of intracranial aneurysms results in significantly lower rates of recurrence compared with BPC, with both coil types showing similar initial occlusion rates and safety profiles.</p>","PeriodicalId":93863,"journal":{"name":"AJNR. American journal of neuroradiology","volume":"46 7","pages":"1379-1386"},"PeriodicalIF":0.0,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144546530","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":"A Review of the Opportunities and Challenges with Large Language Models in Radiology: The Road Ahead.","authors":"Neetu Soni, Manish Ora, Amit Agarwal, Tianbao Yang, Girish Bathla","doi":"10.3174/ajnr.A8589","DOIUrl":"10.3174/ajnr.A8589","url":null,"abstract":"<p><p>In recent years, generative artificial intelligence (AI), particularly large language models (LLMs) and their multimodal counterparts, multimodal large language models, including vision language models, have generated considerable interest in the global AI discourse. LLMs, or pre-trained language models (such as ChatGPT, Med-PaLM, LLaMA), are neural network architectures trained on extensive text data, excelling in language comprehension and generation. Multimodal LLMs, a subset of foundation models, are trained on multimodal data sets, integrating text with another modality, such as images, to learn universal representations akin to human cognition better. This versatility enables them to excel in tasks like chatbots, translation, and creative writing while facilitating knowledge sharing through transfer learning, federated learning, and synthetic data creation. Several of these models can have potentially appealing applications in the medical domain, including, but not limited to, enhancing patient care by processing patient data; summarizing reports and relevant literature; providing diagnostic, treatment, and follow-up recommendations; and ancillary tasks like coding and billing. As radiologists enter this promising but uncharted territory, it is imperative for them to be familiar with the basic terminology and processes of LLMs. Herein, we present an overview of the LLMs and their potential applications and challenges in the imaging domain.</p>","PeriodicalId":93863,"journal":{"name":"AJNR. American journal of neuroradiology","volume":" ","pages":"1292-1299"},"PeriodicalIF":0.0,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142689924","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":"Cross-Sectional Validation of an Automated Lesion Segmentation Software in Multiple Sclerosis: Comparison with Radiologist Assessments.","authors":"Maria Vittoria Spampinato, Heather R Collins, Hannah Wells, William Dennis, Jordan H Chamberlin, Emily Ye, Justin A Chetta, Maria Gisele Matheus, Seth T Stalcup, Donna R Roberts","doi":"10.3174/ajnr.A8655","DOIUrl":"10.3174/ajnr.A8655","url":null,"abstract":"<p><strong>Background and purpose: </strong>MRI is widely used to assess disease burden in MS. This study aimed to evaluate the effectiveness of a commercially available k-nearest neighbors (kNN) network software in quantifying white matter lesion (WML) burden in MS. We compared the software's WML quantification to expert radiologists' assessments.</p><p><strong>Materials and methods: </strong>We retrospectively reviewed brain MRI examinations of adult patients with MS and of adult patients without MS and with a normal brain MRI referred from the neurology clinic. MR images were processed by using an AI-powered, cloud-based kNN software, which generated a DICOM lesion distribution map and a report of WML count and volume in 4 brain regions (periventricular, deep, juxtacortical, and infratentorial white matter). Two blinded radiologists performed semiquantitative assessments of WM lesion load and lesion segmentation accuracy. Additionally, 4 blinded neuroradiologists independently reviewed the data to determine if MRI findings supported an MS diagnosis. The associations between radiologist-rated WML load and kNN model WML volume and count were evaluated with Spearman rank order correlation coefficient (rho) because these variables were not normally distributed. Results were considered significant when <i>P</i> < .05.</p><p><strong>Results: </strong>The study included 32 patients with MS (35.4 years ±9.1) and 19 patients without MS (33.5 years ±12.1). The kNN software demonstrated 94.1% and 84.3% accuracy in differentiating MS from non-MS subjects based respectively on WML count and WML volume, compared with radiologists' accuracy of 90.2% to 94.1%. Lesion segmentation was more accurate for the deep WM and infratentorial regions than for the juxtacortical region (both <i>P</i> < .001).</p><p><strong>Conclusions: </strong>kNN-derived WML volume and WML count provide valuable quantitative metrics of disease burden in MS. AI-powered postprocessing software may enhance the interpretation of brain MRIs in MS patients.</p>","PeriodicalId":93863,"journal":{"name":"AJNR. American journal of neuroradiology","volume":" ","pages":"1510-1516"},"PeriodicalIF":0.0,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142980992","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":"Cracking the Code of Calcification: How Presence and Burden among Intracranial Arteries Influence Stroke Incidence and Recurrence.","authors":"Matteo Conte, Mohammed O Alalfi, Riccardo Cau, Roberta Scicolone, Seemant Chaturvedi, Renu Virmani, Gianluca De Rubeis, Daniel Bos, Luca Saba","doi":"10.3174/ajnr.A8668","DOIUrl":"10.3174/ajnr.A8668","url":null,"abstract":"<p><strong>Background: </strong>Intracranial atherosclerosis accounts for approximately 8% of all strokes in Western societies but the influence of arterial calcification on plaque instability is a topic of ongoing debate.</p><p><strong>Purpose: </strong>Our purpose is to explore the association between the presence and burden of calcium in atherosclerotic plaques among intracranial arteries with the risk of clinical or silent stroke events through a systematic review and meta-analysis.</p><p><strong>Data sources: </strong>Studies from PubMed and Embase investigating intracranial vessel calcification and stroke events were reviewed from inception through May 2024, adhering to PRISMA guidelines.</p><p><strong>Study selection: </strong>Eight longitudinal studies involving 7297 adult patients undergoing CT or CTA scans for symptomatic or asymptomatic intracranial atherosclerosis were included.</p><p><strong>Data analysis: </strong>Pooled odds ratios were calculated to assess the relationship between stroke events and either the presence or burden of intracranial arterial calcification. Quality assessment was conducted using QUADAS-2; overall evidence was established using GRADE system. Meta-analysis was performed using random-effects models.</p><p><strong>Data synthesis: </strong>After adjusting for confounding factors, the presence of intracranial arterial calcification was significantly associated with stroke incidence or recurrence (OR = 1.54; 95% CI 1.06-2.24, <i>P</i> < .001). The strength of this association was found to be similar (OR = 1.56; 95% CI 1.11-2.19, <i>P</i> < .001). A positive correlation was also found for calcium burden (OR = 1.31; 95% CI, 1.17-1.46; <i>P</i> < .001). Heterogeneity was moderate for calcium presence (Q = 13.16 and 9.19; I² = 62% and 42.61%, respectively); negligible for burden analysis (Q = 6.01; I² = 0.01%).</p><p><strong>Limitations: </strong>Despite strict inclusion criteria, heterogeneity and variability in calcium scoring methods across studies were observed. The lack of segment-specific analysis may have limited clinical interpretation.</p><p><strong>Conclusions: </strong>This meta-analysis demonstrates a weak yet present association between intracranial arterial calcification and stroke events. However, given the high prevalence of calcification in the general population, its role for stroke prediction has limited evidence. Future studies may focus on specific arterial segments and emerging calcification patterns to improve predictive accuracy.</p>","PeriodicalId":93863,"journal":{"name":"AJNR. American journal of neuroradiology","volume":" ","pages":"1321-1328"},"PeriodicalIF":0.0,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143043904","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":"High-Resolution MR Imaging of the Parasellar Ligaments.","authors":"Ian T Mark, Myung-Ho In, Daehun Kang, Eric Stinson, John Huston, Yunhong Shu, Luciano Leonel, Maria Peris Celda","doi":"10.3174/ajnr.A8658","DOIUrl":"10.3174/ajnr.A8658","url":null,"abstract":"<p><p>The parasellar ligaments have been previously described in cadaver specimens and intraoperatively, but identification on MR imaging has eluded radiologists. Using high-resolution T2-weighted MR imaging, we identified the parasellar ligaments as T2-hypointense, bandlike structures that emanate from the medial wall of the cavernous sinus. Subsequent dissection of the same specimen provided matching anatomic images of the parasellar ligaments identified on MRI. This imaging finding is important because resection of the medial wall of the cavernous sinus has been tied to improved outcomes for gross total resection and endocrinologic remission of functioning pituitary adenomas.</p>","PeriodicalId":93863,"journal":{"name":"AJNR. American journal of neuroradiology","volume":" ","pages":"1318-1320"},"PeriodicalIF":0.0,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144334623","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":"Endovascular Treatment of Wide-Neck Intracranial Aneurysms Using the Novel Contour Neurovascular System: 5-Year Follow-up.","authors":"Howra Ktayen, Christopher Yusuf Akhunbay-Fudge, Atul Kumar Tyagi, Kenan Deniz, Tufail Patankar","doi":"10.3174/ajnr.A8641","DOIUrl":"10.3174/ajnr.A8641","url":null,"abstract":"<p><strong>Background and purpose: </strong>The Contour neurovascular embolization device is a novel way to treat wide-neck bifurcation aneurysms (WNBA), which often pose considerable treatment challenges. In this study, we aim to evaluate the efficacy and safety profile of this device.</p><p><strong>Materials and methods: </strong>Prospective clinical and radiologic data were collected for all patients treated with the Contour device at our center, between January 2017 and December 2018. All patients were treated electively, and aneurysms were unruptured.</p><p><strong>Results: </strong>Fourteen patients were recruited, and the device was successfully deployed in 11 patients. All patients were women with a mean age of 65 years. Four basilar tip, 2 internal carotid, 3 middle cerebral, 1 anterior communicating, and 1 superior cerebellar artery aneurysms were treated. The mean aneurysmal size was 6 mm (width) × 7.6 mm (height), with 4.1 mm neck. Follow-up imaging included DSA, MRA, and CTA. For the 9 patients available at year 2 follow-up, 5 showed improved occlusion class over time with 8 of 9 having adequate occlusion defined by class 1 and 2 of the Raymond-Roy (R-R) classification system. Eight patients were available for 3-year follow-up: 7 patients had stable occlusion class including 3 patients with stable complete occlusion (R-R class 1). One patient had worsening of R-R occlusion class from 1 to 2 and subsequently presented with acute subarachnoid hemorrhage. Follow-up data 5 years after implantation were available for 8 patients (including the ruptured and re-treated patient): 7 patients had adequate occlusion (R-R class 1 and 2), however, 1 patient had worsened from R-R class 1 to R-R class 2. Four patients had complete occlusion.</p><p><strong>Conclusions: </strong>Results demonstrate progressive occlusion of wide-neck aneurysms over the first 2 years, but we have also demonstrated worsening of R-R occlusion class in some aneurysms that were previously completely occluded. Our results suggest that the Contour device is a good option in WNBA, however, it appears that patients must be followed up for a minimum of 5 years.</p>","PeriodicalId":93863,"journal":{"name":"AJNR. American journal of neuroradiology","volume":" ","pages":"1387-1394"},"PeriodicalIF":0.0,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144287441","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":"A Method for Imaging the Ischemic Penumbra with MRI Using Intravoxel Incoherent Motion.","authors":"Mira M Liu, Niloufar Saadat, Steven P Roth, Marek A Niekrasz, Mihai Giurcanu, Mohammed Salman Shazeeb, Timothy J Carroll, Gregory A Christoforidis","doi":"10.3174/ajnr.A8656","DOIUrl":"10.3174/ajnr.A8656","url":null,"abstract":"<p><strong>Background and purpose: </strong>In acute ischemic stroke, the amount of \"local\" CBF distal to the occlusion, ie, all blood flow, whether supplied antegrade or delayed and dispersed through the collateral network, may contain valuable information regarding infarct growth rate and treatment response. DSC processed with a local arterial input function (AIF) is one method of measuring local CBF (local-qCBF) and has been shown to correlate with collateral supply. Similarly, intravoxel incoherent motion MRI (IVIM) is \"local,\" with excitation and readout in the same plane, and a potential alternative way to measure local-qCBF. This work compares IVIM local-qCBF against DSC local-qCBF in the ischemic penumbra, compares the measurement of perfusion-diffusion mismatch (PWI/DWI), and examines if local-qCBF may improve prediction of the final infarct.</p><p><strong>Materials and methods: </strong>Eight experiments in a preclinical canine model of middle cerebral artery occlusion were performed. Native collateral circulation was quantified via x-ray DSA 30 minutes postocclusion, and collateral supply was subsequently enhanced in a subset of experiments with simultaneous pressor and vasodilator. IVIM, DSC, and DWI MRI were acquired 2.5 hours postocclusion. IVIM was postprocessed to return local-qCBF from fD*, water transport time (WTT) from D*, diffusion from D, and the PWI/DWI mismatch. These were compared with DSC parameters processed first with a standard global-AIF and then with a local-AIF. These DSC parameters included time-to-maximum, local MTT, standard-qCBF, local-qCBF, and PWI/DWI mismatch. Infarct volume was measured with DWI at 2.5 hours and 4 hours postocclusion.</p><p><strong>Results: </strong>Two and one-half hours postocclusion, IVIM local-qCBF in the noninfarcted ipsilateral territory correlated strongly with DSC local-qCBF ([Formula: see text]). Correlation was weaker between IVIM local-qCBF and DSC standard-qCBF ([Formula: see text]). DSC local-qCBF and IVIM local-qCBF in the noninfarcted ipsilateral territory both returned strong prediction of final infarct volume ([Formula: see text]). DSC standard-qCBF was a weaker predictor ([Formula: see text]). The hypoperfused lesion from DSC local-qCBF and from IVIM local-qCBF both predicted final infarct volume with good sensitivity and correlation ([Formula: see text]. The IVIM PWI/DWI ratio was correlated with infarct growth ([Formula: see text]), and WTT correlated with DSC MTT ([Formula: see text]).</p><p><strong>Conclusions: </strong>Noncontrast IVIM measurement of local-qCBF and PWI/DWI mismatch may include collateral circulation and improve prediction of infarct growth.</p>","PeriodicalId":93863,"journal":{"name":"AJNR. American journal of neuroradiology","volume":" ","pages":"1336-1344"},"PeriodicalIF":0.0,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142980989","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":"Deep Learning-Based Prediction of PET Amyloid Status Using MRI.","authors":"Donghoon Kim, Jon André Ottesen, Ashwin Kumar, Brandon C Ho, Elsa Bismuth, Christina B Young, Elizabeth Mormino, Greg Zaharchuk","doi":"10.3174/ajnr.A8899","DOIUrl":"https://doi.org/10.3174/ajnr.A8899","url":null,"abstract":"<p><strong>Background and purpose: </strong>Identifying amyloid-beta (Aβ)-positive patients is essential for Alzheimer's disease (AD) clinical trials and disease-modifying treatments but currently requires PET or cerebrospinal fluid sampling. Previous MRI-based deep learning models, using only T1-weighted (T1w) images, have shown moderate performance.</p><p><strong>Materials and methods: </strong>Multi-contrast MRI and PET-based quantitative Aβ deposition were retrospectively obtained from three public datasets: ADNI, OASIS3, and A4. Aβ positivity was defined using each dataset's recommended centiloid threshold. Two EfficientNet models were trained to predict amyloid positivity: one using only T1w images and another incorporating both T1w and T2-FLAIR. Model performance was assessed using an internal held-out test set, evaluating AUC, accuracy, sensitivity, and specificity. External validation was conducted using an independent cohort from Stanford Alzheimer's Disease Research Center. DeLong's and McNemar's tests were used to compare AUC and accuracy, respectively.</p><p><strong>Results: </strong>A total of 4,056 exams (mean [SD] age: 71.6 [6.3] years; 55% female; 55% amyloid-positive) were used for network development, and 149 exams were used for external testing (mean [SD] age: 72.1 [9.6] years; 58% female; 56% amyloid-positive). The multi-contrast model outperformed the single-modality model in the internal held-out test set (AUC: 0.67, 95% CI: 0.65-0.70, <i>P</i> < 0.001; accuracy: 0.63, 95% CI: 0.62-0.65, <i>P</i> < 0.001) compared to the T1w-only model (AUC: 0.61; accuracy: 0.59). Among cognitive subgroups, the highest performance (AUC: 0.71) was observed in mild cognitive impairment. The multi-contrast model also demonstrated consistent performance in the external test set (AUC: 0.65, 95% CI: 0.60-0.71, <i>P</i> = 0.014; accuracy: 0.62, 95% CI: 0.58- 0.65, <i>P</i> < 0.001).</p><p><strong>Conclusions: </strong>The use of multi-contrast MRI, specifically incorporating T2-FLAIR in addition to T1w images, significantly improved the predictive accuracy of PET-determined amyloid status from MRI scans using a deep learning approach.</p><p><strong>Abbreviations: </strong>Aβ＝ amyloid-beta; AD＝ Alzheimer's disease; AUC＝ area under the receiver operating characteristic curve; CN＝ cognitively normal; MCI＝ mild cognitive impairment; T1w = T1-wegithed; T2-FLAIR = T2-weighted fluid attenuated inversion recovery; FBP＝¹⁸F-florbetapir; FBB＝¹⁸F-florbetaben; SUVR＝ standard uptake value ratio.</p>","PeriodicalId":93863,"journal":{"name":"AJNR. American journal of neuroradiology","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144512917","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}