European Journal of Radiology最新文献

{"title":"Image normalization techniques and their effect on the robustness and predictive power of breast MRI radiomics","authors":"Florian Schwarzhans ,&nbsp;Geevarghese George ,&nbsp;Lorena Escudero Sanchez ,&nbsp;Olgica Zaric ,&nbsp;Jean E. Abraham ,&nbsp;Ramona Woitek ,&nbsp;Sepideh Hatamikia","doi":"10.1016/j.ejrad.2025.112086","DOIUrl":"10.1016/j.ejrad.2025.112086","url":null,"abstract":"<div><h3>Background and purpose</h3><div>Radiomics analysis has emerged as a promising approach to aid in cancer diagnosis and treatment. However, radiomics research currently lacks standardization, and radiomics features can be highly dependent on acquisition and pre-processing techniques used. In this study, we aim to investigate the effect of various image normalization techniques on robustness of radiomics features extracted from breast cancer patient MRI scans.</div></div><div><h3>Materials and methods</h3><div>MRI scans from the publicly available MAMA-MIA dataset and an internal breast MRI test set depicting triple negative breast cancer (TNBC) were used. We compared the effect of commonly used image normalization techniques on radiomics feature robustness<!--> <!-->using Concordance-Correlation-Coefficient (CCC) between multiple combinations of normalization approaches. We also trained machine learning-based prediction models of pathologic complete response (pCR) on radiomics after different normalization techniques were used and compared their areas under the receiver operating characteristic curve (ROC-AUC).</div></div><div><h3>Results</h3><div>For predicting complete pathological response from pre-treatment breast cancer MRI radiomics, the highest overall ROC-AUC was achieved by using a combination of three different normalization techniques indicating their potentially powerful role when working with heterogeneous imaging data. The effect of normalization was more pronounced with smaller training data and normalization may be less important with increasing abundance of training data. Additionally, we observed considerable differences between MRI data sets and their feature robustness towards normalization.</div></div><div><h3>Conclusion</h3><div>Overall, we were able to demonstrate the importance of selecting and standardizing normalization methods for accurate and reliable radiomics analysis in breast MRI scans especially with small training data sets.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"187 ","pages":"Article 112086"},"PeriodicalIF":3.2,"publicationDate":"2025-03-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143761125","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Very rare adverse reactions to intravascular contrast media: From Kounis to sweet syndrome","authors":"Aart J. van der Molen ,&nbsp;José J. Laguna ,&nbsp;Annick A.J.M van de Ven ,&nbsp;Francisco Vega","doi":"10.1016/j.ejrad.2025.112066","DOIUrl":"10.1016/j.ejrad.2025.112066","url":null,"abstract":"<div><div>Intravascular administration of contrast media may inadvertently result in adverse drug reactions (ADR) including selected hypersensitivity reactions. This review highlights some very rare ADR, aiming to alert imaging physicians to these conditions and to prompt adequate management. This is particularly relevant for Kounis syndrome, an immediate hypersensitivity reaction involving the coronary arteries. The acute management of Kounis syndrome can be challenging, as it requires continued contrast media administration, while balancing the simultaneous coronary reperfusion that requires vasodilatation with anti-allergic treatment that may involve vasoconstrictor agents. For other adverse reactions, referral to a drug allergy specialist is highly recommended to assess causality and identify safe alternatives.</div></div><div><h3>Clinical Relevance Statement</h3><div>Cardiologists and radiologists must recognize the key symptoms of these very rare adverse reactions to contrast media to promptly initiate the appropriate therapy and refer patients for allergological evaluation, ideally within 1–6 months after the reaction.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"187 ","pages":"Article 112066"},"PeriodicalIF":3.2,"publicationDate":"2025-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143739907","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The age and gender-dependent changes in pulmonary arterial flow distribution: A cardiac magnetic resonance flow study","authors":"Gülbanu Güner,&nbsp;Safiye Sanem Dereli Bulut","doi":"10.1016/j.ejrad.2025.112064","DOIUrl":"10.1016/j.ejrad.2025.112064","url":null,"abstract":"<div><h3>Objective</h3><div>To investigate whether the pulmonary artery blood flow distribution in the normal population changes depending on age and gender using the Cardiac Magnetic Resonance Phase Contrast Imaging method.</div></div><div><h3>Materials and methods</h3><div>A total of 62 healthy volunteers aged 20–60 years were included in the study between September 2020 and November 2021. Participants were first divided into two groups according to their age (group 120–40 years old, group 2: 40–60 years old). Then, each group was divided into two according to gender (groups 1A, 1B and groups 2A, 2B). After routine MR sequences of mediastinum and heart (True Fast Imaging with Steady State Free Precession, short axis for heart (SA), 4-chamber, 2-chamber and 3-chamber CINE sequence). Phase Contrast imaging was performed to Main Pulmonary Artery (MPA), right PA and left PA. On the phase contrast (PC) images obtained, flow velocity, forward flow, backward flow and net forward flow volume and gradient were calculated for each artery. With these data, the distribution ratio for both pulmonary arteries was calculated. Measurements were made simultaneously by two radiologists<strong>.</strong></div></div><div><h3>Results</h3><div>In phase contrast sequence evaluation; No statistically significant correlation was found between RPA/LPA distribution ratio and Group 1 (p &gt; 0.05). A moderate and positive correlation was found statistically between RPA/LPA distribution ratio and Group 2 (r:0.42, p:0.04). No statistically significant correlation was found between the RPA/LPA distribution ratio and the gender variable (p:0.59).</div></div><div><h3>Conclusion</h3><div>The distribution ratio of pulmonary blood flow to both lungs in healthy volunteers can be measured using PC-CMR without the need for invasive procedures. Knowing the normal distribution ratios according to age in the population can be used for the early diagnosis of pathological conditions.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"187 ","pages":"Article 112064"},"PeriodicalIF":3.2,"publicationDate":"2025-03-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143715208","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Prediction of postoperative pancreatic fistula based on multi-sequence MR imaging","authors":"Xiaoyang Li ,&nbsp;Kuinan Tong ,&nbsp;Tianxin Cheng ,&nbsp;Piao Yan ,&nbsp;Hui Xu ,&nbsp;Kun Liu ,&nbsp;Rui Xu ,&nbsp;Jun Lu ,&nbsp;Zhenghan Yang ,&nbsp;Hongwei Wu","doi":"10.1016/j.ejrad.2025.112067","DOIUrl":"10.1016/j.ejrad.2025.112067","url":null,"abstract":"<div><h3>Objective</h3><div>To investigate the value of conventional MRI sequences in predicting the occurrence of postoperative pancreatic fistula (POPF) in patients undergoing pancreaticoduodenectomy (PD).</div></div><div><h3>Methods</h3><div>A total of 122 patients from August 2019 to April 2023 were enrolled. All patients underwent pancreatic histological evaluation, including fibrosis, fat deposition, and acinar cell atrophy. The preoperative image features of pancreas were obtained, including morphological features, pancreas-muscle signal intensity ratio, pancreatic fat fraction and multi-phase enhancement features. The patients were divided into two groups according to whether pancreatic fistula occurred after operation. The related risk factors of pancreatic fistula, the correlation between imaging and pathological changes were analyzed, and the value of preoperative imaging in predicting pancreatic fistula was evaluated.</div></div><div><h3>Results</h3><div>Of the 122 patients, 23(18.9 %) developed POPF. Pathological score showed that there was a significant difference in pancreatic fat deposition between the two groups (P = 0.006), the fat deposition score was higher in the POPF group. Pancreatic fat deposition was the only independent risk factor for POPF(OR,1.933; P = 0.018). MRI showed that proton density fat fraction(PDFF) (P = 0.001), pancreas-to-aorta signal intensity ratio(P-A SI ratio) of equilibrium phase(P = 0.023) and delay phase(P = 0.020) had significant differences. PDFF was positively correlated with fat deposition(r = 0.404, P &lt; 0.001), P-A SI ratio of equilibrium phase and delay phase were positively correlated with fibrosis(r = 0.313, P = 0.002; r = 0.315, P = 0.002, respectively). ROC analysis showed that PDFF had the best efficacy in predicting postoperative pancreatic fistula (AUROC = 0.810), better than P-A SI ratio of equilibrium phase(AUROC = 0.752) and delayed phase(AUROC = 0.766).</div></div><div><h3>Conclusions</h3><div>Pancreatic fat deposition is a high risk factor for POPF, PDFF can reflect fat deposition and predict POPF.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"186 ","pages":"Article 112067"},"PeriodicalIF":3.2,"publicationDate":"2025-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143697973","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Magnetic resonance imaging assessment of the discoid glenoid labrum","authors":"Kirthi Sathyakumar ,&nbsp;Praveen Kumar Chinniah ,&nbsp;Madhavi Kandagaddala ,&nbsp;Anil Oommen ,&nbsp;Chandrasekaran ,&nbsp;Jeremy Bliss","doi":"10.1016/j.ejrad.2025.112065","DOIUrl":"10.1016/j.ejrad.2025.112065","url":null,"abstract":"<div><h3>Background</h3><div>The glenoid labrum varies in size and shape, sometimes resembling the knee’s meniscus with a free inner edge projecting into the joint. This incomplete discoid-shaped or meniscoid glenoid labrum is a seldom recognized rare variant.</div></div><div><h3>Purpose</h3><div>To determine the prevalence and magnetic resonance imaging [MRI] characteristics of discoid glenoid labrum.</div></div><div><h3>Material and methods</h3><div>Retrospective review of the MRI shoulder joints of 201 patients. The MR images were evaluated for the presence of a discoid labrum, and various characteristics were assessed using predefined criteria.</div></div><div><h3>Results</h3><div>Among the 201 patients, a discoid glenoid labrum was identified in 27 cases; incidence rate of 13.4 %. The mean age of these patients was 41.85 years [range 15–73 years, SD ± 16.9]. The most common position was anterior or antero-inferior [n = 25; 92.6 %], with the apices of almost all discoid folds found between 3–6o’clock position. Twenty-one [77.8 %] of the discoid folds were small, and 6 [22.2 %] were large. Tears or degeneration were observed in 9 [33.3 %]. A statistically significant difference between genders was found in age [p = 0.026] and width [p = 0.044], with discoid labrum appearing in younger men and covering a larger part of the glenoid fossa. The labra were widest at the 3o’clock position [p = 0.008]. Additionally, we discovered a novel radiological finding termed the ’reversed Oreo cookie sign’ in discoid glenoid labra.</div></div><div><h3>Conclusion</h3><div>The discoid glenoid labrum is a shoulder variant that is often overlooked. Recognizing this condition is crucial to avoid misdiagnosing labral pathology and to understand its potential susceptibility to labral tears, degeneration, and shoulder pain.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"187 ","pages":"Article 112065"},"PeriodicalIF":3.2,"publicationDate":"2025-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143706314","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Optimization of shoulder synthetic MRI through post-processing and comparison with conventional MRI","authors":"Naoto Nishizawa ,&nbsp;Hidetake Yabuuchi ,&nbsp;Kei Nishikawa ,&nbsp;Tatsuhiro Wada ,&nbsp;Kouji Kobayashi ,&nbsp;Chiaki Tokunaga ,&nbsp;Tsukasa Kojima ,&nbsp;Takumi Ohnishi ,&nbsp;Yuji Yano ,&nbsp;Kouji Sagiyama ,&nbsp;Tomoyuki Hida ,&nbsp;Yuzo Yamasaki ,&nbsp;Takuya Hino ,&nbsp;Kousei Ishigami","doi":"10.1016/j.ejrad.2025.112069","DOIUrl":"10.1016/j.ejrad.2025.112069","url":null,"abstract":"<div><h3>Purpose</h3><div>To evaluate the utility of synthetic MRI of the shoulder compared with conventional MRI and to optimize the parameters of morphological images through post-processing. Additionally, we aimed to assess the agreement of T2 values between multi-echo spin-echo (MESE) and multi-dynamic multi-echo (MDME) sequences.</div></div><div><h3>Methods</h3><div>Twenty healthy volunteers underwent shoulder MRI. The optimal post-processing parameters for the tendon–muscle contrast were examined using synthetic MRI, and two radiologists analyzed three image sets: conventional images, synthetic images using preset parameters, and optimized images. Qualitative analysis included assessment of the visibility of anatomical structures, overall image quality, and magic angle effect, whereas the quantitative analysis included measurement of the relative signal intensity and relative contrast. The T2 values of the articular cartilage and supraspinatus muscle were measured for each sequence.</div></div><div><h3>Results</h3><div>Images synthesized with short echo times and long repetition times showed high tendon–muscle contrast. For fat-suppressed T2-weighted images, conventional images showed the highest image quality (p &lt; 0.001), whereas the optimized images achieved comparable visibility of the rotator cuff (p = 0.031–1.0). No significant differences were observed among image sets in proton density-weighted images and T1-weighted images (p &gt; 0.05). The T2 values of the MDME sequence were consistent with those of the MESE sequence at the muscle (p = 0.81), but were approximately 8.3 ms longer at the cartilage (p &lt; 0.001).</div></div><div><h3>Conclusions</h3><div>Synthetic MRI provided acceptable image quality using appropriate post-processing parameters. The simultaneous acquisition of multiple morphological images and quantitative maps within five minutes holds promise for shoulder examination.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"186 ","pages":"Article 112069"},"PeriodicalIF":3.2,"publicationDate":"2025-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143714235","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Predicting hepatocellular carcinoma response to TACE: A machine learning study based on 2.5D CT imaging and deep features analysis","authors":"Chong Lin ,&nbsp;Ting Cao ,&nbsp;Maowen Tang ,&nbsp;Wei Pu ,&nbsp;Pinggui Lei","doi":"10.1016/j.ejrad.2025.112060","DOIUrl":"10.1016/j.ejrad.2025.112060","url":null,"abstract":"<div><h3>Objectives</h3><div>Prior to the commencement of treatment, it is essential to establish an objective method for accurately predicting the prognosis of patients with hepatocellular carcinoma (HCC) undergoing transarterial chemoembolization (TACE). In this study, we aimed to develop a machine learning (ML) model to predict the response of HCC patients to TACE based on CT images analysis.</div></div><div><h3>Materials and methods</h3><div>Public dataset from The Cancer Imaging Archive (TCIA), uploaded in August 2022, comprised a total of 105 cases, including 68 males and 37 females. The external testing dataset was collected from March 1, 2019 to July 1, 2022, consisting of total of 26 patients who underwent TACE treatment at our institution and were followed up for at least 3 months after TACE, including 22 males and 4 females. The public dataset was utilized for ResNet50 transfer learning and ML model construction, while the external testing dataset was used for model performance evaluation. All CT images with the largest lesions in axial, sagittal, and coronal orientations were selected to construct 2.5D images. Pre-trained ResNet50 weights were adapted through transfer learning to serve as a feature extractor to derive deep features for building ML models. Model performance was assessed using area under the curve (AUC), accuracy, F1-Score, confusion matrix analysis, decision curves, and calibration curves.</div></div><div><h3>Results</h3><div>The AUC values for the external testing dataset were 0.90, 0.90, 0.91, and 0.89 for random forest classifier (RFC), support vector classifier (SVC), logistic regression (LR), and extreme gradient boosting (XGB), respectively. The accuracy values for the external testing dataset were 0.79, 0.81, 0.80, and 0.80 for RFC, SVC, LR, and XGB, respectively. The F1-score values for the external testing dataset were 0.75, 0.77, 0.78, and 0.79 for RFC, SVC, LR, and XGB, respectively.</div></div><div><h3>Conclusion</h3><div>The ML model constructed using deep features from 2.5D images has the potential to be applied in predicting the prognosis of HCC patients following TACE treatment.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"187 ","pages":"Article 112060"},"PeriodicalIF":3.2,"publicationDate":"2025-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143734620","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Assessment of fetal hepatic lipid content by magnetic resonance imaging and association of results with clinical maternal and fetal parameters","authors":"Vanessa Berger-Kulemann ,&nbsp;Daniela Prayer ,&nbsp;Nina Sieberer ,&nbsp;Gregor Kasprian ,&nbsp;Gregor Dovjak ,&nbsp;Jürgen Harreiter ,&nbsp;Alexandra Kautzky-Willer ,&nbsp;Michael Weber ,&nbsp;Martin Krššák ,&nbsp;Anke Scharrer ,&nbsp;Marlene Stuempflen","doi":"10.1016/j.ejrad.2025.112061","DOIUrl":"10.1016/j.ejrad.2025.112061","url":null,"abstract":"<div><h3>Purpose</h3><div>To quantify hepatocellular lipids in the fetal liver, we tested the feasibility of the multiecho mDixon Quant sequence (chemical shift encoded magnetic resonance imaging (MRI)) during clinically routine fetal whole-body MRI and investigated the correlation of hepatocellular lipids with different clinical maternal and fetal parameters.</div></div><div><h3>Methods</h3><div>The livers of 155 fetuses were prospectively investigated with multiecho CSE-MRI sequences during clinically indicated whole-body MRI, performed between gestational weeks 19 and 38 on a 1.5 Tesla scanner. The hepatocellular lipids were quantified by measuring the proton density fat fraction in the left and right liver lobe. Results of the right liver lobe were correlated with the maternal body mass index, maternal age, presence of maternal diabetes, gestational age at assessment, estimated fetal weight, fetal sex, and birth weight.</div></div><div><h3>Results</h3><div>Quantification of fetal hepatocellular lipids was feasible in 151/155 (97.4 %) fetuses. Four examinations were excluded due to strong motion artifacts and poor image quality. The proton density fat fraction values ranged from 0 % to 5.7 % (mean 2.26; SD 1.37). Hepatocellular lipids were associated with the presence of maternal diabetes (p = 0.027). No association was found between hepatocellular lipids and maternal body mass index (p = 0.306), maternal age (p = 0.582), gestational age (p = 0.456), estimated fetal weight (p = 0.176), fetal sex (p = 0.181), or birth weight (p = 0.957).</div></div><div><h3>Conclusion</h3><div>Quantification of fetal hepatocellular lipids is feasible and may routinely be performed during whole-body MRI to detect early liver fat accumulation, particularly in the presence of maternal diabetes.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"186 ","pages":"Article 112061"},"PeriodicalIF":3.2,"publicationDate":"2025-03-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143696677","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Carotid plaque CTA analysis in patients with initial and recurrent ischemic stroke","authors":"Hui Zhao ,&nbsp;Yanqi Shan ,&nbsp;Wei Hu ,&nbsp;Moran Yang ,&nbsp;Xijun Gong ,&nbsp;Qisong Zhou ,&nbsp;Dai Zhang ,&nbsp;Jing Jin","doi":"10.1016/j.ejrad.2025.112058","DOIUrl":"10.1016/j.ejrad.2025.112058","url":null,"abstract":"&lt;div&gt;&lt;h3&gt;Objectives&lt;/h3&gt;&lt;div&gt;Recurrent ischemic stroke leads to prolonged hospital stays, deterioration in functional outcomes, increased disability and mortality rates. The composition of plaques is closely related to the occurrence of stroke. Our aim was to explore differences in carotid plaque characteristics between patients with initial and recurrent ischemic stroke.&lt;/div&gt;&lt;/div&gt;&lt;div&gt;&lt;h3&gt;Methods&lt;/h3&gt;&lt;div&gt;A retrospective study was conducted on 208 anterior circulation stroke patients who had completed head and neck CT angiography (CTA) examinations. Patients were divided into initial cerebral infarction (ICI) group and recurrent cerebral infarction (RCI) group according to clinical and imaging data. The histological and morphological characteristics of the plaques were assessed by using VascuCAP software. The general demographic data, plaque characteristics, and serological indicators were compared between the two groups. Then, the association between plaque characteristics were investigated. To investigate the contribution of plaque characteristics for predicting the risk of stroke recurrence, we have constructed multivariate logistic regression models with plaque characteristics and blood lipid level parameters.&lt;/div&gt;&lt;/div&gt;&lt;div&gt;&lt;h3&gt;Results&lt;/h3&gt;&lt;div&gt;There were significant differences in total carotid plaque volume (&lt;em&gt;P&lt;/em&gt; = 0.002) between the ICI and RCI groups. The volume and percentage of Intraplaque Hemorrhage (IPH) and calcification (CALC) in the RCI group were significantly higher than those in the ICI group (&lt;em&gt;P&lt;/em&gt; &lt; 0.05). There were also statistically significant differences in IPH volume (P=＜0.001)and IPH percentage (P＜0.001)between the symptomatic and asymptomatic sides in the RCI group. In the ICI group, the difference in the maximum stenosis of the carotid artery lumen was statistically significant between the 2 sides (P = 0.025). The total cholesterol (&lt;em&gt;P&lt;/em&gt; = 0.024) and triglyceride levels (&lt;em&gt;P&lt;/em&gt; = 0.049) was significantly larger in the ICI group than in the RCI group. Further, there is a significant negative correlation between the volumes and percentages of IPH and CALC in the ICI group (volume, &lt;em&gt;r&lt;/em&gt; = -0.225, &lt;em&gt;P&lt;/em&gt; = 0.021; percentage, &lt;em&gt;r&lt;/em&gt; = -0.293, &lt;em&gt;P&lt;/em&gt; = 0.002). In addition, IPH volume and percentage were independent predictors in the logistic model with combined features of plaque volume, plaque percentage and blood lipid level (AUC = 0.755, 95 %CI 0.690–0.820).&lt;/div&gt;&lt;/div&gt;&lt;div&gt;&lt;h3&gt;Conclusion&lt;/h3&gt;&lt;div&gt;The volume of total carotid plaque and IPH were significantly different between the ICI group and RCI group. The IPH volume and percentage on the symptomatic side of the RCI group were higher than those on the asymptomatic side. IPH has important indicative significance in identifying patients at risk of recurrent stroke.&lt;/div&gt;&lt;div&gt;Abbreviations: CTA, computed tomography angiography; ICI, initial cerebral infarction; RCI, recurrent cerebral infarction; IPH, intraplaque hemorrhage; ","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"186 ","pages":"Article 112058"},"PeriodicalIF":3.2,"publicationDate":"2025-03-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143687795","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Machine learning models for enhanced diagnosis and risk assessment of prostate cancer with 68Ga-PSMA-617 PET/CT","authors":"Wenhao Zhu ,&nbsp;Yongxiang Tang ,&nbsp;Lin Qi ,&nbsp;Xiaomei Gao ,&nbsp;Shuo Hu ,&nbsp;Min-Feng Chen ,&nbsp;Yi Cai","doi":"10.1016/j.ejrad.2025.112063","DOIUrl":"10.1016/j.ejrad.2025.112063","url":null,"abstract":"<div><h3>Objective</h3><div>Prostate cancer (PCa) is highly heterogeneous, making early detection of adverse pathological features crucial for improving patient outcomes. This study aims to predict PCa aggressiveness and identify radiomic and protein biomarkers associated with poor pathology, ultimately developing a multi-omics marker model for better clinical risk stratification.</div></div><div><h3>Methods</h3><div>In this retrospective study, 191 patients with PCa or benign prostatic hyperplasia confirmed via ⁶⁸Ga-PSMA-617 PET/CT scans were analyzed. Radiomic features were extracted from scan contours, and six machine learning algorithms were used to predict malignancy and adverse pathological features like Gleason score, ISUP group, tumor stage, lymph node infiltration, and perineural invasion. Feature selection and dimensionality reduction were performed using minimum redundancy maximum relevance and least absolute shrinkage and selection operator methods. Proteomics analysis on 39 patients identified protein biomarkers, followed by correlation analysis between radiomic features and identified proteins.</div></div><div><h3>Results</h3><div>The radiomics model showed an AUC of 0.938 for predicting malignant prostate lesions and 0.916 for adverse pathological features in the test set, with validation set AUCs of 0.918 and 0.855, respectively. Three quantitative radiomic features and ten protein molecules associated with adverse pathology were identified, with significant correlations observed between radiomic features and protein biomarkers. Radioproteomic analysis revealed that molecular changes in protein molecules could influence imaging biomarkers.</div></div><div><h3>Conclusions</h3><div>The machine learning models based on 68 Ga-PSMA-617 PET/CT radiomic features performed well in stratifying patients, supporting clinical risk stratification and highlighting connections between radiomic characteristics and protein biomarkers.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"186 ","pages":"Article 112063"},"PeriodicalIF":3.2,"publicationDate":"2025-03-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143697974","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}