European Journal of Radiology Open最新文献

{"title":"Multidisciplinary quantitative and qualitative assessment of IDH-mutant gliomas with full diagnostic deep learning image reconstruction.","authors":"Christer Ruff, Paula Bombach, Constantin Roder, Eliane Weinbrenner, Christoph Artzner, Leonie Zerweck, Frank Paulsen, Till-Karsten Hauser, Ulrike Ernemann, Georg Gohla","doi":"10.1016/j.ejro.2024.100617","DOIUrl":"10.1016/j.ejro.2024.100617","url":null,"abstract":"<p><p><b>Rationale and Objectives:</b> Diagnostic accuracy and therapeutic decision-making for IDH-mutant gliomas in tumor board reviews are based on MRI and multidisciplinary interactions.</p><p><strong>Materials and methods: </strong>This study explores the feasibility of deep learning-based reconstruction (DLR) in MRI for IDH-mutant gliomas. The research utilizes a multidisciplinary approach, engaging neuroradiologists, neurosurgeons, neuro-oncologists, and radiotherapists to evaluate qualitative aspects of DLR and conventional reconstructed (CR) sequences. Furthermore, quantitative image quality and tumor volumes according to Response Assessment in Neuro-Oncology (RANO) 2.0 standards were assessed.</p><p><strong>Results: </strong>All DLR sequences consistently outperformed CR sequences (median of 4 for all) in qualitative image quality across all raters (p < 0.001 for all) and revealed higher SNR and CNR values (p < 0.001 for all). Preference for all DLR over CR was overwhelming, with ratings of 84 % from the neuroradiologist, 100 % from the neurosurgeon, 92 % from the neuro-oncologist, and 84 % from the radiation oncologist. The RANO 2.0 compliant measurements showed no significant difference between the CR and DRL sequences (p = 0.142).</p><p><strong>Conclusion: </strong>This study demonstrates the clinical feasibility of DLR in MR imaging of IDH-mutant gliomas, with significant time savings of 29.6 % on average and non-inferior image quality to CR. DLR sequences received strong multidisciplinary preference, underscoring their potential for enhancing neuro-oncological decision-making and suitability for clinical implementation.</p>","PeriodicalId":38076,"journal":{"name":"European Journal of Radiology Open","volume":"13 ","pages":"100617"},"PeriodicalIF":1.8,"publicationDate":"2024-12-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11664152/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142883233","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Role of pre-procedure CCTA in predicting failed percutaneous coronary intervention for chronic total occlusions","authors":"Hua Zhou ,&nbsp;Xiaojun Fan ,&nbsp;Mingyuan Yuan ,&nbsp;Wei Wang,&nbsp;Qiyuan Wu","doi":"10.1016/j.ejro.2024.100616","DOIUrl":"10.1016/j.ejro.2024.100616","url":null,"abstract":"<div><h3>Purpose</h3><div>This study aimed to identify major lesion characteristics of chronic total occlusions (CTOs) that predict failed percutaneous coronary intervention (PCI) using pre-procedure coronary computed tomography angiography (CCTA) in combination with conventional coronary angiography (CCA).</div></div><div><h3>Methods</h3><div>Consecutive patients with at least one CTO of the native coronary arteries received CCTA and CCA-guided PCI, with computed tomography performed before or during PCI.</div></div><div><h3>Results</h3><div>A total of 76 patients with CTO were included in this study. 76 patients were divided into successful and failed PCI groups based on their PCI outcome. There were 62 (81.58 %) patients in the successful PCI group and 14 (18.42 %) in the failed PCI group. The occlusion length ≥20 mm, ostial or bifurcation lesions, negative remodeling, microchannels, and good collateral vessels were the CCTA morphologic parameters associated with PCI outcome (P&lt;0.05). In addition, the blunt stump, occlusion length ≥20 mm, and ostial or bifurcation lesions, were the CCA morphologic parameters associated with PCI outcome (P&lt;0.05). The multivariate regression model showed that the three independent negative predictors: blunt stump on CCA (OR: 0.63; 95 % CI: 0.23–0.98; p =0.048), occlusion length ≥20 mm on CCTA (OR: 0.37; 95 % CI: 0.32–0.71; p &lt;0.001) and negative remodeling on CCTA (OR: 0.26; 95 % CI: 0.28–0.44; p &lt;0.001).</div></div><div><h3>Conclusion</h3><div>Our study demonstrated that combining CCTA and CCA morphologic characteristics could improve PCI outcome prediction in patients with CTO compared to CCTA morphologic features alone.</div></div>","PeriodicalId":38076,"journal":{"name":"European Journal of Radiology Open","volume":"13 ","pages":"Article 100616"},"PeriodicalIF":1.8,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142746687","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhancing mortality prediction in patients with spontaneous intracerebral hemorrhage: Radiomics and supervised machine learning on non-contrast computed tomography","authors":"Antonio López-Rueda ,&nbsp;María-Ángeles Rodríguez-Sánchez ,&nbsp;Elena Serrano ,&nbsp;Javier Moreno ,&nbsp;Alejandro Rodríguez ,&nbsp;Laura Llull ,&nbsp;Sergi Amaro ,&nbsp;Laura Oleaga","doi":"10.1016/j.ejro.2024.100618","DOIUrl":"10.1016/j.ejro.2024.100618","url":null,"abstract":"<div><h3>Purpose</h3><div>This study aims to develop a Radiomics-based Supervised Machine-Learning model to predict mortality in patients with spontaneous intracerebral hemorrhage (sICH).</div></div><div><h3>Methods</h3><div>Retrospective analysis of a prospectively collected clinical registry of patients with sICH consecutively admitted at a single academic comprehensive stroke center between January-2016 and April-2018. We conducted an in-depth analysis of 105 radiomic features extracted from 105 patients. Following the identification and handling of missing values, radiomics values were scaled to 0–1 to train different classifiers. The sample was split into 80–20 % training-test and validation cohort in a stratified fashion. Random Forest(RF), K-Nearest Neighbor(KNN), and Support Vector Machine(SVM) classifiers were evaluated, along with several feature selection methods and hyperparameter optimization strategies, to classify the binary outcome of mortality or survival during hospital admission. A tenfold stratified cross-validation method was used to train the models, and average metrics were calculated.</div></div><div><h3>Results</h3><div>RF, KNN, and SVM, with the \"DropOut+SelectKBest\" feature selection strategy and no hyperparameter optimization, demonstrated the best performances with the least number of radiomic features and the most simplified models, achieving a sensitivity range between 0.90 and 0.95 and AUC range from 0.97 to 1 on the validation dataset. Regarding the confusion matrix, the SVM model did not predict any false negative test (negative predicted value 1).</div></div><div><h3>Conclusion</h3><div>Radiomics-based Supervised Machine Learning models can predict mortality during admission in patients with sICH. SVM with the \"DropOut+SelectKBest\" feature selection strategy and no hyperparameter optimization was the best simplified model to detect mortality during admission in patients with sICH.</div></div>","PeriodicalId":38076,"journal":{"name":"European Journal of Radiology Open","volume":"13 ","pages":"Article 100618"},"PeriodicalIF":1.8,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142756636","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Different radiomics models in predicting the malignant potential of small intestinal stromal tumors","authors":"Yuxin Xie ,&nbsp;Chongfeng Duan ,&nbsp;Xuzhe Zhou ,&nbsp;Xiaoming Zhou ,&nbsp;Qiulin Shao ,&nbsp;Xin Wang ,&nbsp;Shuai Zhang ,&nbsp;Fang Liu ,&nbsp;Zhenbo Sun ,&nbsp;Ruirui Zhao ,&nbsp;Gang Wang","doi":"10.1016/j.ejro.2024.100615","DOIUrl":"10.1016/j.ejro.2024.100615","url":null,"abstract":"<div><h3>Objectives</h3><div>To explore the feasibility of different radiomics models for predicting the malignant potential of small intestinal stromal tumors (SISTs), and to select the best radiomics model.</div></div><div><h3>Methods</h3><div>A retrospective analysis of 140 patients with SISTs was conducted. Radiomics features were extracted from CT-enhanced images. Support vector machine (SVM), Decision tree (DT), Conditional inference trees (CIT), Random Forest (RF), K-nearest neighbors (KNN), Back-propagation neural network (BPNet), and Bayes were used to construct different radiomics models. The clinical data and CT performance were selected using univariate analysis and to construct clinical model. Nomogram model was developed by combining clinical data and radiomics features. Model performances were assessed by using the area under the receiver operator characteristic (ROC) curve (AUC). The models’ clinical values were assessed by decision curve analysis (DCA).</div></div><div><h3>Results</h3><div>A total of 1132 radiomics features were extracted. Among radiomics models, SVM was better than DT, CIT, RF, KNN, BPNet, Bayes because it had the highest AUC with a significant difference (P&lt;0.05). The AUC of the clinical model was 0.781. The AUC of the radiomics model was 0.910. The AUC of nomogram model was 0.938. Clinical models had the lowest AUC. Nomogram AUC were slightly higher than radiomics model, but the difference was not significant (P=0.48). The DCA of the nomogram model and radiomics model showed optimal clinical efficacy.</div></div><div><h3>Conclusions</h3><div>The model constructed with SVM method was the best model for predicting the malignant potential of SISTs. Radiomics model and nomogram model showed high predictive value in predicting the malignant potential of SISTs.</div></div>","PeriodicalId":38076,"journal":{"name":"European Journal of Radiology Open","volume":"13 ","pages":"Article 100615"},"PeriodicalIF":1.8,"publicationDate":"2024-11-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142701825","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Low-dose lung CT: Optimizing diagnostic radiation dose – A phantom study","authors":"Michael Zellner ,&nbsp;Sebastian Tschauner ,&nbsp;Mathias S. Weyland ,&nbsp;Peter Eggenberger Hotz ,&nbsp;Stephan Scheidegger ,&nbsp;Christian J. Kellenberger","doi":"10.1016/j.ejro.2024.100614","DOIUrl":"10.1016/j.ejro.2024.100614","url":null,"abstract":"<div><h3>Background/purpose</h3><div>To investigate a quantitative method for assessing image quality of low dose lung computed tomography (CT) and find the lowest exposure dose providing diagnostic images.</div></div><div><h3>Methods</h3><div>Axial volumetric lung CT acquisitions (256 slice scanner) were performed on three different sized anthropomorphic phantoms at different dose levels. The maximum steepness of sigmoid curves fitted to line density profiles was measured at lung-to-pleura interfaces. For each phantom, image sharpness was calculated as the median of 468 measurements from 39 different locations. Diagnostic image quality for the adult and paediatric phantom was rated by three radiologists using 4-point Likert scales. The image sharpness cut-off for obtaining adequate image quality was determined from qualitative ratings.</div></div><div><h3>Results</h3><div>Adequate diagnostic image quality was reached at a median steepness of 713 HU/mm in the adult phantom with a corresponding CTDIvol of 0.14 mGy and an effective dose of 0.13 mSv at a dose level of 100 kVp and 10 mA. In the paediatric phantom diagnostic image quality was reached at a median steepness of 1139 HU/mm with a corresponding CTDIvol of 0.13 mGy and an effective dose of 0.08 mSv at a dose level of 100 kVp and 10 mA.</div></div><div><h3>Conclusions</h3><div>Determination of image sharpness on line density profiles can be used as quantitative measure for image quality of lung CT. Sufficient-quality lung CT can be achieved at effective radiation doses of 0.13 mSv (adult phantom) and 0.08 mSv (paediatric phantom). These findings suggest that substantial dose reduction is feasible without compromising diagnostic accuracy.</div></div>","PeriodicalId":38076,"journal":{"name":"European Journal of Radiology Open","volume":"13 ","pages":"Article 100614"},"PeriodicalIF":1.8,"publicationDate":"2024-11-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142701824","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Initial experience with Double-vein Embolization in Hungary","authors":"David Adam Korda ,&nbsp;Andras Bibok ,&nbsp;Attila Doros ,&nbsp;Denes Horvathy ,&nbsp;Oszkar Hahn ,&nbsp;Balint Kokas ,&nbsp;Damjan Pekli ,&nbsp;Anna Zsofia Meltzer ,&nbsp;Attila Szijarto ,&nbsp;Domonkos Nadasdy-Horvath ,&nbsp;Pal Akos Deak","doi":"10.1016/j.ejro.2024.100613","DOIUrl":"10.1016/j.ejro.2024.100613","url":null,"abstract":"<div><h3>Introduction</h3><div>In recent years several new techniques have emerged to induce hypertrophy of the future liver remnant prior to major hepatectomies. We aimed to summarize our initial experience with Double-vein Embolization as the first center in Hungary.</div></div><div><h3>Methods</h3><div>Between March 2023 and August 2024 a total of 16 Double-vein Embolization procedures were performed in Semmelweis University. Future liver remnant volume was calculated based on computed tomography scans obtained within 4 weeks prior and 2–3 weeks after the procedure. Tc-99m mebrofenin hepatobiliary scintigraphy results were available for 12/16 patients.</div></div><div><h3>Results</h3><div>Technical success rate was 100 %. No major complication was observed. Successful resection rate was 93.8 %. One patient died due to post-hepatectomy liver failure. Future liver remnant volume and ratio increased significantly after the procedure compared to baseline (433.1 ± 163.8 cm³ vs. 603.5 ± 201.8 cm³, p &lt; 0.0001 and 27.2 ± 6.5 % vs. 37 ± 8.8 %, p &lt; 0.0001, respectively). Future liver remnant clearance improved significantly 1 and 2 weeks after the procedure (1.68 ± 0.58 %/min/m² vs. 2.44 ± 0.64 %/min/m² and 2.39 ± 0.31 %/min/m², respectively). Mean function gain was 50.6 % after one week and 60.1% after two weeks, respectively.</div></div><div><h3>Discussion</h3><div>Volumetric and functional outcomes in the present study are comparable with results reported in the literature. Our findings provide further evidence that Double-vein Embolization is a safe procedure that offers sufficient volumetric and functional gain in most candidates for liver resection. However, further studies are needed to define the exact place of this new technique in clinical practice.</div></div>","PeriodicalId":38076,"journal":{"name":"European Journal of Radiology Open","volume":"13 ","pages":"Article 100613"},"PeriodicalIF":1.8,"publicationDate":"2024-11-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142701823","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Coronary CT angiography: First comparison of model-based and hybrid iterative reconstruction with the reference standard invasive catheter angiography for CAD-RADS reporting","authors":"Aiste Matuleviciute-Stojanoska ,&nbsp;Julia Sautier ,&nbsp;Verena Bauer ,&nbsp;Martin Nuessel ,&nbsp;Volha Nizhnikava ,&nbsp;Christian Stumpf ,&nbsp;Thorsten Klink","doi":"10.1016/j.ejro.2024.100612","DOIUrl":"10.1016/j.ejro.2024.100612","url":null,"abstract":"<div><h3>Background</h3><div>The purpose of this study was to compare CCTA images generated using HIR and IMR algorithm with the reference standard ICA, and to determine to what extend further improvements of IMR over HIR can be expected.</div></div><div><h3>Methods</h3><div>This retrospective study included 60 patients with low to intermediate CAD risk, who underwent coronary CTA (with HIR and IMR) and ICA. ICA was used as reference standard. Two independent and blinded readers evaluated 2226 segments, classifying stenosis with CAD-RADS (significant stenosis ≥3). Image quality was assessed with a 5-point scale, SNR in the ascending aorta, and FWHM of proximal LCA calibers. The impact of image noise, radiation dose, and BMI on diagnostic accuracy was evaluated using ROC curves and Fisher’s Exact Test. Quantitative plaque analysis was performed on 28 plaques.</div></div><div><h3>Results</h3><div>IMR showed higher image quality than HIR (IMR 4.4, HIR 3.97, p&lt;0.001) with better SNR (21.4 vs. 13.28, p&lt;0.001) and FWHM (4.44 vs. 4.55, p=0.003). IMR had better diagnostic accuracy (ROC AUC 0.967 vs. 0.948, p=0.16, performed better at higher radiation doses (p=0.02) and showed a larger minimum lumen area (p=0.022 and p=0.046).</div></div><div><h3>Conclusion</h3><div>IMR offers significantly superior image quality of CCTA, more precise measurements, and a stronger positive correlation with ICA. The overall diagnostic accuracy may be superior with IMR, although the differences were not statistically significant. However, in patients who are exposed to higher radiation doses during CCTA due to their constitution, IMR enables significantly better diagnostic accuracy than HIR thus providing a specific benefit for obese patients.</div></div>","PeriodicalId":38076,"journal":{"name":"European Journal of Radiology Open","volume":"13 ","pages":"Article 100612"},"PeriodicalIF":1.8,"publicationDate":"2024-11-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142701732","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Increased background parenchymal enhancement on peri-menopausal breast magnetic resonance imaging","authors":"Hidemi Okuma ,&nbsp;Amro Masarwah ,&nbsp;Aleksandr Istomin ,&nbsp;Aki Nykänen ,&nbsp;Juhana Hakumäki ,&nbsp;Ritva Vanninen ,&nbsp;Mazen Sudah","doi":"10.1016/j.ejro.2024.100611","DOIUrl":"10.1016/j.ejro.2024.100611","url":null,"abstract":"<div><h3>Objectives</h3><div>To examine the background parenchymal enhancement (BPE) levels in peri-menopausal breast MRI compared with pre- and post-menopausal breast MRI.</div></div><div><h3>Methods</h3><div>This study included 562 patients (55.8±12.3 years) who underwent contrast-enhanced dynamic breast MRI between 2011 and 2015 for clinical indications. We evaluated the BPE level, amount of fibroglandular tissue (FGT), and social and clinical variables. The inter-reader agreement for the amount of FGT and the BPE level was evaluated using interclass correlation coefficients. Associations between the BPE level and body mass index (BMI), ages of menarche and menopause, childbirth history, number of children, and the amount of FGT were determined using Spearman’s correlation coefficients or Mann-Whitney <em>U</em>-test. Pearson’s χ² test was used to assess the difference in the frequency of BPE categories among the age-groups.</div></div><div><h3>Results</h3><div>The inter-reader agreement was 0.864 for the amount of FGT and 0.840 for the BPE level, both indicating almost perfect agreement. The BPE level showed a weak positive correlation with the amount of FGT (Spearman’s ρ=0.271, <em>P</em>&lt;0.001). BPE was not significantly correlated with BMI, childbirth history, number of births, or ages of menarche or menopause. BPE was greater in the peri-menopausal age-group compared with the corresponding pre- and post-menopausal age-groups, both with benign and malignant lesions.</div></div><div><h3>Conclusions</h3><div>BPE was greater in the peri-menopausal stage than in the pre- and post-menopausal stages. Our results suggest that BPE showed a non-linear decrease with age and that the hormonal disbalance in the peri-menopausal period has a greater effect on the BPE level than was previously assumed.</div></div>","PeriodicalId":38076,"journal":{"name":"European Journal of Radiology Open","volume":"13 ","pages":"Article 100611"},"PeriodicalIF":1.8,"publicationDate":"2024-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142701761","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep learning based on multiparametric MRI predicts early recurrence in hepatocellular carcinoma patients with solitary tumors ≤5 cm","authors":"Tingting Mu ,&nbsp;Xinde Zheng ,&nbsp;Danjun Song ,&nbsp;Jiejun Chen ,&nbsp;Xuewang Yue ,&nbsp;Wentao Wang ,&nbsp;Shengxiang Rao","doi":"10.1016/j.ejro.2024.100610","DOIUrl":"10.1016/j.ejro.2024.100610","url":null,"abstract":"<div><h3>Purpose</h3><div>To evaluate the effectiveness of a constructed deep learning model in predicting early recurrence after surgery in hepatocellular carcinoma (HCC) patients with solitary tumors ≤5 cm.</div></div><div><h3>Materials and methods</h3><div>Our study included a total of 331 HCC patients who underwent curative resection, with all patients having preoperative dynamic contrast-enhanced MRI (DCE-MRI). Patients who recurred within two years after surgery were defined as early recurrence. The enrolled patients were randomly divided into the training group and the testing group. A ResNet-based deep learning model with eight conventional neural network branches was built to predict the early recurrence status of these patients. Patient characteristics and laboratory tests were further filtered by regression models and then integrated with deep learning models to improve the prediction performance.</div></div><div><h3>Results</h3><div>Among 331 HCC patients, 70 (21.1 %) experienced early recurrence. In multivariate Cox regression analysis, only tumor size (Hazard ratio (HR=1.394, 95 %CI:1.011–1.920, p value=0.043) and deep learning extracted image features (HR: 38440, 95 %CI:2321–636600, p value&lt;0.001) were significant risk factors for early recurrence. In the training and testing cohort, the AUCs of the image-based deep learning prediction model were 0.839 and 0.833. By integrating tumor size with image-based deep learning model to construct a combined model, we found that the AUCs of the combined model to assess early recurrence in the training and validation cohort were 0.846 and 0.842. We further developed a nomogram to visualize the preoperative combined model, and the prediction performance of nomogram showed a good fitness in the testing cohort.</div></div><div><h3>Conclusions</h3><div>The proposed deep learning-based prediction model using DCE-MRI is useful for assessing early recurrence in HCC patients with single tumors ≤5 cm.</div></div>","PeriodicalId":38076,"journal":{"name":"European Journal of Radiology Open","volume":"13 ","pages":"Article 100610"},"PeriodicalIF":1.8,"publicationDate":"2024-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142652863","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep learning model for diagnosis of thyroid nodules with size less than 1 cm: A multicenter, retrospective study","authors":"Na Feng ,&nbsp;Shanshan Zhao ,&nbsp;Kai Wang ,&nbsp;Peizhe Chen ,&nbsp;Yunpeng Wang ,&nbsp;Yuan Gao ,&nbsp;Zhengping Wang ,&nbsp;Yidan Lu ,&nbsp;Chen Chen ,&nbsp;Jincao Yao ,&nbsp;Zhikai Lei ,&nbsp;Dong Xu","doi":"10.1016/j.ejro.2024.100609","DOIUrl":"10.1016/j.ejro.2024.100609","url":null,"abstract":"<div><h3>Objective</h3><div>To develop a ultrasound images based dual-channel deep learning model to achieve accurate early diagnosis of thyroid nodules less than 1 cm.</div></div><div><h3>Methods</h3><div>A dual-channel deep learning model called thyroid nodule transformer network (TNT-Net) was proposed. The model has two input channels for transverse and longitudinal ultrasound images of thyroid nodules, respectively. A total of 9649 nodules from 8455 patients across five hospitals were retrospectively collected. The data were divided into a training set (8453 nodules, 7369 patients), an internal test set (565 nodules, 512 patients), and an external test set (631 nodules, 574 patients).</div></div><div><h3>Results</h3><div>TNT-Net achieved an area under the curve (AUC) of 0.953 (95 % confidence interval (CI): 0.934, 0.969) on the internal test set and 0.941 (95 % CI: 0.921, 0.957) on the external test set, significantly outperforming traditional deep convolutional neural network models and single-channel swin transformer model, whose AUCs ranged from 0.800 (95 % CI: 0.759, 0.837) to 0.856 (95 % CI: 0.819, 0.881). Furthermore, feature heatmap visualization showed that TNT-Net could extract richer and more energetic malignant nodule patterns.</div></div><div><h3>Conclusion</h3><div>The proposed TNT-Net model significantly improved the recognition capability for thyroid nodules with size less than 1 cm. This model has the potential to reduce overdiagnosis and overtreatment of such nodules, providing essential support for precise management of thyroid nodules while complementing fine-needle aspiration biopsy.</div></div>","PeriodicalId":38076,"journal":{"name":"European Journal of Radiology Open","volume":"13 ","pages":"Article 100609"},"PeriodicalIF":1.8,"publicationDate":"2024-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142561079","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}