The aim of this study is to explore the utility of Inductive Decision Tree models (IDTs) in distinguishing between benign, malignant, and high-risk (B3) breast lesions.
Materials and Methods
We analyzed 124 histologically confirmed lesions in 114 patients who underwent breast MR with BI-RADS 4 or 5 assessment. The dataset comprised 10 imaging parameters and one clinical observation. Using the IDTs method (algorithm C5.0 boosted with AdaBoost algorithm) combined with the data balancing method SMOTE (Synthetic Minority Oversampling Technique) and a corresponding new method called LCC (Leveling of Cases per Class), we developed corresponding 3-class classification models (Benign, B3, or Malignant). The training set used for classification model development consists of 112 cases with 12 variables, and the model’s performance was assessed using 10-fold Cross-Validation and Leave-One-Out methods (utilizing the training set) and the Use Test Set method (testing on an unknown (for the models) dataset of 12 cases with 12 variables).
Results
This preliminary study demonstrates the potential for IDTs to accurately distinguish between benign, B3 and Malignant lesions based on extracted data from breast MRI exams with a high classification accuracy (88.70 %), mean sensitivity of 97.18 % and specificity of 98.59 % achieved by the optimal classification model, derived from the combination of the IDTs method and the LCC data balancing method.
{"title":"Accuracy of distinguishing benign, high-risk lesions and malignancies with inductive machine learning models in BIRADS 4 and BIRADS 5 lesions on breast MR examinations","authors":"Evangelia Panourgias , Evangelos Karampotsis , Natalia Douma , Charis Bourgioti , Vassilis Koutoulidis , George Rigas , Lia Moulopoulos , Georgios Dounias","doi":"10.1016/j.ejrad.2024.111801","DOIUrl":"10.1016/j.ejrad.2024.111801","url":null,"abstract":"<div><h3>Rationale and Objectives</h3><div>The aim of this study is to explore the utility of Inductive Decision Tree models (IDTs) in distinguishing between benign, malignant, and high-risk (B3) breast lesions.</div></div><div><h3>Materials and Methods</h3><div>We analyzed 124 histologically confirmed lesions in 114 patients who underwent breast MR with BI-RADS 4 or 5 assessment. The dataset comprised 10 imaging parameters and one clinical observation. Using the IDTs method (algorithm C5.0 boosted with AdaBoost algorithm) combined with the data balancing method SMOTE (Synthetic Minority Oversampling Technique) and a corresponding new method called LCC (Leveling of Cases per Class), we developed corresponding 3-class classification models (Benign, B3, or Malignant). The training set used for classification model development consists of 112 cases with 12 variables, and the model’s performance was assessed using 10-fold Cross-Validation and Leave-One-Out methods (utilizing the training set) and the Use Test Set method (testing on an unknown (for the models) dataset of 12 cases with 12 variables).</div></div><div><h3>Results</h3><div>This preliminary study demonstrates the potential for IDTs to accurately distinguish between benign, B3 and Malignant lesions based on extracted data from breast MRI exams with a high classification accuracy (88.70 %), mean sensitivity of 97.18 % and specificity of 98.59 % achieved by the optimal classification model, derived from the combination of the IDTs method and the LCC data balancing method.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"181 ","pages":"Article 111801"},"PeriodicalIF":3.2,"publicationDate":"2024-10-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142603771","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}
Pub Date : 2024-10-24DOI: 10.1016/j.ejrad.2024.111802
Iva Petkovska , Or Alus , Lee Rodriguez , Maria El Homsi , Jennifer S Golia Pernicka , Maria Clara Fernandes , Junting Zheng , Marinela Capanu , Ricardo Otazo
Background
To evaluate the effectiveness of a deep learning denoising approach to accelerate diffusion-weighted imaging (DWI) and thus improve diagnostic accuracy and image quality in restaging rectal MRI following total neoadjuvant therapy (TNT).
Methods
This retrospective single-center study included patients with locally advanced rectal cancer who underwent restaging rectal MRI between December 30, 2021, and June 1, 2022, following TNT. A convolutional neural network trained with DWI data was employed to denoise accelerated DWI acquisitions (i.e., acquisitions performed with a reduced number of repetitions compared to standard acquisitions). Image characteristics and residual disease were independently assessed by two radiologists across original and denoised images. Statistical analyses included the Wilcoxon signed-rank test to compare image quality scores across denoised and original images, weighted kappa statistics for inter-reader agreement assessment, and the calculation of measures of diagnostic accuracy.
Results
In 46 patients (median age, 60 years [IQR: 47–72]; 37 men and 9 women), 8- and 16-fold accelerated images maintained or exhibited enhanced lesion visibility and image quality compared with original images that were performed 16 repetitions. Denoised images maintained diagnostic accuracy, with conditional specificities of up to 96 %. Moderate-to-high inter-reader agreement indicated reliable image and diagnostic assessment. The overall test yield for denoised DWI reconstructions ranged from 76–98 %, demonstrating a reduction in equivocal interpretations.
Conclusion
Applying a denoising network to accelerate rectal DWI acquisitions can reduce scan times and enhance image quality while maintaining diagnostic accuracy, presenting a potential pathway for more efficient rectal cancer management.
{"title":"Clinical evaluation of accelerated diffusion-weighted imaging of rectal cancer using a denoising neural network","authors":"Iva Petkovska , Or Alus , Lee Rodriguez , Maria El Homsi , Jennifer S Golia Pernicka , Maria Clara Fernandes , Junting Zheng , Marinela Capanu , Ricardo Otazo","doi":"10.1016/j.ejrad.2024.111802","DOIUrl":"10.1016/j.ejrad.2024.111802","url":null,"abstract":"<div><h3>Background</h3><div>To evaluate the effectiveness of a deep learning denoising approach to accelerate diffusion-weighted imaging (DWI) and thus improve diagnostic accuracy and image quality in restaging rectal MRI following total neoadjuvant therapy (TNT).</div></div><div><h3>Methods</h3><div>This retrospective single-center study included patients with locally advanced rectal cancer who underwent restaging rectal MRI between December 30, 2021, and June 1, 2022, following TNT. A convolutional neural network trained with DWI data was employed to denoise accelerated DWI acquisitions (i.e., acquisitions performed with a reduced number of repetitions compared to standard acquisitions). Image characteristics and residual disease were independently assessed by two radiologists across original and denoised images. Statistical analyses included the Wilcoxon signed-rank test to compare image quality scores across denoised and original images, weighted kappa statistics for inter-reader agreement assessment, and the calculation of measures of diagnostic accuracy.</div></div><div><h3>Results</h3><div>In 46 patients (median age, 60 years [IQR: 47–72]; 37 men and 9 women), 8- and 16-fold accelerated images maintained or exhibited enhanced lesion visibility and image quality compared with original images that were performed 16 repetitions. Denoised images maintained diagnostic accuracy, with conditional specificities of up to 96 %. Moderate-to-high inter-reader agreement indicated reliable image and diagnostic assessment. The overall test yield for denoised DWI reconstructions ranged from 76–98 %, demonstrating a reduction in equivocal interpretations.</div></div><div><h3>Conclusion</h3><div>Applying a denoising network to accelerate rectal DWI acquisitions can reduce scan times and enhance image quality while maintaining diagnostic accuracy, presenting a potential pathway for more efficient rectal cancer management.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"181 ","pages":"Article 111802"},"PeriodicalIF":3.2,"publicationDate":"2024-10-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142521374","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}
Pub Date : 2024-10-24DOI: 10.1016/j.ejrad.2024.111803
Aart J. van der Molen , Francisco Vega , Annick A.J.M van de Ven , Ilona A. Dekkers , José J. Laguna
The risk of hypersensitivity reactions (HSR) following nonvascular administration of contrast media (CM) for diagnostic studies is very low, likely due to minimal absorption into the systemic circulation. Most published individual cases of HSR after nonvascular CM administration are immediate reactions caused by ionic high-osmolar CM, few by nonionic low-osmolar CM, and none by gadolinium-based contrast agents. Measures to prevent recurrent HSR following nonvascular administration are similar to those recommended to prevent HSR after intravascular CM administration. Premedication as preventive measure has been abandoned, while switching to an alternative CM, preferably based on the results of an allergological analysis, is increasingly advocated. In selected scenarios, preventive measures may be minimized.
在诊断研究中使用造影剂（CM）进行非血管给药后发生超敏反应（HSR）的风险非常低，这可能是由于进入全身循环的吸收量极少。已发表的非血管性使用造影剂后发生超敏反应的个案中，大多数是由离子型高渗透压造影剂引起的即刻反应，少数是由非离子型低渗透压造影剂引起的，而钆类造影剂则没有引起超敏反应。预防非血管给药后 HSR 复发的措施与预防血管内 CM 给药后 HSR 的措施类似。目前已放弃将预先用药作为预防措施，而越来越多的人主张改用其他 CM，最好是根据过敏学分析的结果。在某些情况下，可尽量减少预防措施。
{"title":"Hypersensitivity reactions after diagnostic nonvascular administration of iodine-based contrast media and gadolinium-based contrast agents and the role of the drug allergy specialist","authors":"Aart J. van der Molen , Francisco Vega , Annick A.J.M van de Ven , Ilona A. Dekkers , José J. Laguna","doi":"10.1016/j.ejrad.2024.111803","DOIUrl":"10.1016/j.ejrad.2024.111803","url":null,"abstract":"<div><div>The risk of hypersensitivity reactions (HSR) following nonvascular administration of contrast media (CM) for diagnostic studies is very low, likely due to minimal absorption into the systemic circulation. Most published individual cases of HSR after nonvascular CM administration are immediate reactions caused by ionic high-osmolar CM, few by nonionic low-osmolar CM, and none by gadolinium-based contrast agents. Measures to prevent recurrent HSR following nonvascular administration are similar to those recommended to prevent HSR after intravascular CM administration. Premedication as preventive measure has been abandoned, while switching to an alternative CM, preferably based on the results of an allergological analysis, is increasingly advocated. In selected scenarios, preventive measures may be minimized.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"181 ","pages":"Article 111803"},"PeriodicalIF":3.2,"publicationDate":"2024-10-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142521375","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}
To compare the adherence of the interpretation and reporting staging system, respectively proposed in the 2012 and 2016 European Society of Gastrointestinal and Abdominal Radiology (ESGAR) Guidelines for Magnetic Resonance Imaging (MRI) staging of rectal cancer, focusing on the improvement offered by the criteria introduced by 2016 ESGAR guidelines.
Method
Fifty-six patients affected by rectal cancer were included; 25/56 patients underwent upfront surgery; 31 underwent to neo-adjuvant chemo-radiotherapy before surgery. All patients underwent 3 T MRI examination for local staging. All MR exams were evaluated by two radiologists with 20- and 4-years’ experience, who were blinded to each other; the T and N stages, the Mesorectal Fascia (MRF) status and the Extramural Vascular Invasion (EMVI) were assessed according to both 2012 and 2016 ESGAR guidelines. The correlation between radiological and pathological findings, as well as the MRI staging were evaluated.
Results
As to the expert reviewer, no significant differences were found by comparing the MR T and N stages, T and N restaging, MRF status and EMVI according to 2012 and 2016 ESGAR guidelines. As to the 4-years’ experience radiologist the MR staging agreement between 2012 and 2016 guidelines was “moderate” in N-stage evaluation and “fair” in T-restaging evaluation. No significant discrepancies were found for other parameters.
Conclusions
MRI is a reliable method in rectal cancer staging/restaging. The assessment of T-restaging is improved by adopting the 2016 ESGAR guidelines, especially for non-expert readers; this result could be justified by the introduction of diffusion-weighted imaging. On the contrary, the newest guidelines do not improve the diagnostic performance in assessing nodal staging and restaging.
目的 比较2012年和2016年欧洲胃肠道和腹部放射学会（ESGAR）《直肠癌磁共振成像（MRI）分期指南》分别提出的解释和报告分期系统的遵循情况，重点关注2016年ESGAR指南引入的标准所带来的改进。方法 纳入56例直肠癌患者；25/56例患者接受了前期手术；31例患者在手术前接受了新辅助化疗和放疗。所有患者均接受了 3 T MRI 检查，以进行局部分期。所有磁共振检查均由两位分别有20年和4年经验的放射科医生进行评估，他们互不设盲；T期和N期、中直肠筋膜（MRF）状态和壁外血管侵犯（EMVI）均根据2012年和2016年ESGAR指南进行评估。结果就专家评审员而言，根据 2012 年和 2016 年 ESGAR 指南比较 MR T 和 N 分期、T 和 N 重分期、MRF 状态和 EMVI，未发现显著差异。对于有 4 年经验的放射科医生而言，2012 年和 2016 年指南在 N 期评估方面的 MR 分期一致性为 "中度"，而在 T 重分期评估方面为 "尚可"。结论 MRI 是直肠癌分期/预后的可靠方法。采用2016年ESGAR指南后，T分期评估得到了改善，尤其是对非专业读者而言；这一结果可能是由于引入了弥散加权成像。相反，最新指南并没有提高评估结节分期和再分期的诊断性能。
{"title":"MR staging of rectal cancer: Comparison between the 2012 and 2016 European Society of Gastrointestinal and Abdominal Radiology (ESGAR) Guidelines","authors":"Piero Boraschi , Francescamaria Donati , Rosa Cervelli , Kathrine Bani , Riccardo Morganti , Niccolò Furbetta , Luca Morelli , Emanuele Neri","doi":"10.1016/j.ejrad.2024.111804","DOIUrl":"10.1016/j.ejrad.2024.111804","url":null,"abstract":"<div><h3>Purpose</h3><div>To compare the adherence of the interpretation and reporting staging system, respectively proposed in the 2012 and 2016 European Society of Gastrointestinal and Abdominal Radiology (ESGAR) Guidelines for Magnetic Resonance Imaging (MRI) staging of rectal cancer, focusing on the improvement offered by the criteria introduced by 2016 ESGAR guidelines.</div></div><div><h3>Method</h3><div>Fifty-six patients affected by rectal cancer were included; 25/56 patients underwent upfront surgery; 31 underwent to neo-adjuvant chemo-radiotherapy before surgery. All patients underwent 3 T MRI examination for local staging. All MR exams were evaluated by two radiologists with 20- and 4-years’ experience, who were blinded to each other; the T and N stages, the Mesorectal Fascia (MRF) status and the Extramural Vascular Invasion (EMVI) were assessed according to both 2012 and 2016 ESGAR guidelines. The correlation between radiological and pathological findings, as well as the MRI staging were evaluated.</div></div><div><h3>Results</h3><div>As to the expert reviewer, no significant differences were found by comparing the MR T and N stages, T and N restaging, MRF status and EMVI according to 2012 and 2016 ESGAR guidelines. As to the 4-years’ experience radiologist the MR staging agreement between 2012 and 2016 guidelines was “moderate” in N-stage evaluation and “fair” in T-restaging evaluation. No significant discrepancies were found for other parameters.</div></div><div><h3>Conclusions</h3><div>MRI is a reliable method in rectal cancer staging/restaging. The assessment of T-restaging is improved by adopting the 2016 ESGAR guidelines, especially for non-expert readers; this result could be justified by the introduction of diffusion-weighted imaging. On the contrary, the newest guidelines do not improve the diagnostic performance in assessing nodal staging and restaging.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"181 ","pages":"Article 111804"},"PeriodicalIF":3.2,"publicationDate":"2024-10-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142537853","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}
To identify patients with atypical ductal hyperplasia (ADH) at low risk of upgrading to carcinoma. This study aims to assess the performance of radiomics combined with clinical factors to predict occult breast cancer among women diagnosed with ADH.
Methods
This study retrospectively included microcalcification clusters of patients who underwent Mx and VABB with a diagnosis of ADH at a tertiary center from January 2015 to May 2023. Clinical and radiological data (age, cluster size, BI-RADS classification, mammographic density, breast cancer history, residual microcalcifications) were collected. Surgical outcomes were used to determine upgrade. Four logistic regression models were developed to predict the risk of upgrade. The performance was evaluated using the area under the receiver operating characteristic curve (AUC) and performance scores.
Results
A total of 143 patients with 153 clusters were included. Twelve radiomic features and six clinical factors were selected for model development. The sample was split into 107 training and 46 test cases. Clinical features achieved an AUC of 0.72 (0.60–0.84), radiomic features an AUC of 0.73 (0.61–0.85). Radiomic features with “cluster size” and “age” improved the AUC to 0.79 (0.67–0.91). The best model, incorporating all data, achieved an AUC of 0.82 (0.71–0.92), a specificity of 0.89 (0.75, 0.97), and NPV of 0.92 (0.78–0.98).
Conclusion
This study demonstrates the potential of radiomic as a valuable tool for reducing unnecessary treatments for patient classified as “low risk of ADH upgrade”. Combining radiomic information with clinical data improved the accuracy of risk prediction.
{"title":"Radiomic and clinical model for predicting atypical ductal hyperplasia upgrades and potentially reduce unnecessary surgical treatments","authors":"Nicole Brunetti , Cristina Campi , Giorgia Biddau , Michele Piana , Ilaria Picone , Benedetta Conti , Sara Cesano , Oleksandr Starovatskyi , Silvia Bozzano , Giuseppe Rescinito , Simona Tosto , Alessandro Garlaschi , Massimo Calabrese , Alberto Stefano Tagliafico","doi":"10.1016/j.ejrad.2024.111799","DOIUrl":"10.1016/j.ejrad.2024.111799","url":null,"abstract":"<div><h3>Objective</h3><div>To identify patients with atypical ductal hyperplasia (ADH) at low risk of upgrading to carcinoma. This study aims to assess the performance of radiomics combined with clinical factors to predict occult breast cancer among women diagnosed with ADH.</div></div><div><h3>Methods</h3><div>This study retrospectively included microcalcification clusters of patients who underwent Mx and VABB with a diagnosis of ADH at a tertiary center from January 2015 to May 2023. Clinical and radiological data (age, cluster size, BI-RADS classification, mammographic density, breast cancer history, residual microcalcifications) were collected. Surgical outcomes were used to determine upgrade. Four logistic regression models were developed to predict the risk of upgrade. The performance was evaluated using the area under the receiver operating characteristic curve (AUC) and performance scores.</div></div><div><h3>Results</h3><div>A total of 143 patients with 153 clusters were included. Twelve radiomic features and six clinical factors were selected for model development. The sample was split into 107 training and 46 test cases. Clinical features achieved an AUC of 0.72 (0.60–0.84), radiomic features an AUC of 0.73 (0.61–0.85). Radiomic features with “cluster size” and “age” improved the AUC to 0.79 (0.67–0.91). The best model, incorporating all data, achieved an AUC of 0.82 (0.71–0.92), a specificity of 0.89 (0.75, 0.97), and NPV of 0.92 (0.78–0.98).</div></div><div><h3>Conclusion</h3><div>This study demonstrates the potential of radiomic as a valuable tool for reducing unnecessary treatments for patient classified as “low risk of ADH upgrade”. Combining radiomic information with clinical data improved the accuracy of risk prediction.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"181 ","pages":"Article 111799"},"PeriodicalIF":3.2,"publicationDate":"2024-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142497620","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}
Pub Date : 2024-10-22DOI: 10.1016/j.ejrad.2024.111800
Wanling Li , Leihao Sha , Jiayu Zhu , Fan Long , Lei Chen
Objective
Epilepsy is the most common complication and cause of morbidity and mortality in tuberous sclerosis complex (TSC). Surgery is associated with an increased probability of achieving seizure-freedom. However, the preoperative noninvasive localisation of epileptogenic tubers remains challenging. This study aimed to identify multimodal magnetic resonance imaging (MRI) biomarkers of epilepsy in patients with TSC and develop a prediction model of epileptogenicity in these patients.
Methods
Patients with TSC, with or without epilepsy, were recruited. All patients underwent MRI scanning, including T1WI, T2WI, T2W-FLAIR, DTI, and multi-parametric MR with a flexible design (MULTIPLEX). We compared the multimodal cerebral MRI characteristics of the cortical tubers, subependymal nodules, and perilesional tissue between patients with TSC with or without epilepsy and developed a prediction model for epileptogenicity.
Results
Among the patients with TSC, 32 with and 16 without epilepsy underwent MRI. Higher proton-density mapping (PD) of cortical tubers and decreased fractional anisotropy (FA), increased mean diffusivity (MD), and increased radial diffusivity (RD) of subependymal nodules were associated with epileptogenicity in both the centre and perilesional tissue, independent of TSC gene variation. Based on the above findings, we developed a prediction model for epileptogenicity with an area under the curve of 0.973, specificity of 0.909, and sensitivity of 0.963 (P < 0.001).
Conclusion
In patients with TSC, high PD of the cortical tubers, decreased FA, and elevated MD/RD of the subependymal nodules were significantly associated with epileptogenicity. A prediction model based on multimodal cerebral MRI characteristics has the potential to evaluate the likelihood of epilepsy in patients with TSC.
{"title":"Prediction of epileptogenicity in patients with tuberous sclerosis complex using multimodal cerebral MRI","authors":"Wanling Li , Leihao Sha , Jiayu Zhu , Fan Long , Lei Chen","doi":"10.1016/j.ejrad.2024.111800","DOIUrl":"10.1016/j.ejrad.2024.111800","url":null,"abstract":"<div><h3>Objective</h3><div>Epilepsy is the most common complication and cause of morbidity and mortality in tuberous sclerosis complex (TSC). Surgery is associated with an increased probability of achieving seizure-freedom. However, the preoperative noninvasive localisation of epileptogenic tubers remains challenging. This study aimed to identify multimodal magnetic resonance imaging (MRI) biomarkers of epilepsy in patients with TSC and develop a prediction model of epileptogenicity in these patients.</div></div><div><h3>Methods</h3><div>Patients with TSC, with or without epilepsy, were recruited. All patients underwent MRI scanning, including T1WI, T2WI, T2W-FLAIR, DTI, and multi-parametric MR with a flexible design (MULTIPLEX). We compared the multimodal cerebral MRI characteristics of the cortical tubers, subependymal nodules, and perilesional tissue between patients with TSC with or without epilepsy and developed a prediction model for epileptogenicity.</div></div><div><h3>Results</h3><div>Among the patients with TSC, 32 with and 16 without epilepsy underwent MRI. Higher proton-density mapping (PD) of cortical tubers and decreased fractional anisotropy (FA), increased mean diffusivity (MD), and increased radial diffusivity (RD) of subependymal nodules were associated with epileptogenicity in both the centre and perilesional tissue, independent of TSC gene variation. Based on the above findings, we developed a prediction model for epileptogenicity with an area under the curve of 0.973, specificity of 0.909, and sensitivity of 0.963 (P < 0.001).</div></div><div><h3>Conclusion</h3><div>In patients with TSC, high PD of the cortical tubers, decreased FA, and elevated MD/RD of the subependymal nodules were significantly associated with epileptogenicity. A prediction model based on multimodal cerebral MRI characteristics has the potential to evaluate the likelihood of epilepsy in patients with TSC.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"181 ","pages":"Article 111800"},"PeriodicalIF":3.2,"publicationDate":"2024-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142497618","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}
Pub Date : 2024-10-21DOI: 10.1016/j.ejrad.2024.111796
Hayes Pearce , Yu-Cherng Chang , Marcia C. Javitt , Jashodeep Datta , Agustin Pimentel , Steven Bialick , Peter J. Hosein , Francesco Alessandrino
Liquid biopsy with sequencing of circulating tumor DNA (ctDNA) is a minimally invasive method for sampling body fluids and offers a promising alternative to tissue biopsies that involve greater risks, costs, and time. ctDNA not only identifies actionable targets by revealing unique molecular signatures in cancer, but also may assess treatment response, treatment resistance and progression, and recurrence. Imaging correlates of these applications are already being identified and utilized for various solid tumors.
Radiologists have new challenges in interpreting oncologic imaging. Given their integral role in cancer surveillance, they must become familiar with the importance of ctDNA in detecting recurrence and minimal residual disease, measuring treatment response, predicting survival and metastatic patterns, and identifying new molecular therapeutic targets.
In this review, we provide an overview of ctDNA testing, and a snapshot of current clinical guidelines from the National Comprehensive Cancer Network and the European Society of Molecular Oncology on the use of ctDNA in lung, breast, colorectal, pancreatic, and hepatobiliary cancers. For each cancer type, we also highlight current research applications of ctDNA that are relevant to the field of diagnostic radiology.
循环肿瘤DNA（ctDNA）测序的液体活检是一种微创的体液采样方法，有望替代风险更大、成本更高、时间更长的组织活检。这些应用的相关成像技术已被确定并用于各种实体瘤。放射医师在解读肿瘤成像方面面临着新的挑战。鉴于他们在癌症监控中不可或缺的作用，他们必须熟悉ctDNA 在检测复发和最小残留病、衡量治疗反应、预测生存和转移模式以及确定新的分子治疗靶点方面的重要性。在本综述中，我们将概述 ctDNA 检测，并简要介绍美国国家综合癌症网络（National Comprehensive Cancer Network）和欧洲分子肿瘤学会（European Society of Molecular Oncology）关于在肺癌、乳腺癌、结直肠癌、胰腺癌和肝胆癌中使用 ctDNA 的现行临床指南。针对每种癌症类型，我们还重点介绍了目前与放射诊断领域相关的 ctDNA 研究应用。
{"title":"ctDNA in the reading room: A guide for radiologists","authors":"Hayes Pearce , Yu-Cherng Chang , Marcia C. Javitt , Jashodeep Datta , Agustin Pimentel , Steven Bialick , Peter J. Hosein , Francesco Alessandrino","doi":"10.1016/j.ejrad.2024.111796","DOIUrl":"10.1016/j.ejrad.2024.111796","url":null,"abstract":"<div><div>Liquid biopsy with sequencing of circulating tumor DNA (ctDNA) is a minimally invasive method for sampling body fluids and offers a promising alternative to tissue biopsies that involve greater risks, costs, and time. ctDNA not only identifies actionable targets by revealing unique molecular signatures in cancer, but also may assess treatment response, treatment resistance and progression, and recurrence. Imaging correlates of these applications are already being identified and utilized for various solid tumors.</div><div>Radiologists have new challenges in interpreting oncologic imaging. Given their integral role in cancer surveillance, they must become familiar with the importance of ctDNA in detecting recurrence and minimal residual disease, measuring treatment response, predicting survival and metastatic patterns, and identifying new molecular therapeutic targets.</div><div>In this review, we provide an overview of ctDNA testing, and a snapshot of current clinical guidelines from the National Comprehensive Cancer Network and the European Society of Molecular Oncology on the use of ctDNA in lung, breast, colorectal, pancreatic, and hepatobiliary cancers. For each cancer type, we also highlight current research applications of ctDNA that are relevant to the field of diagnostic radiology.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"181 ","pages":"Article 111796"},"PeriodicalIF":3.2,"publicationDate":"2024-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142497613","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}
Pub Date : 2024-10-20DOI: 10.1016/j.ejrad.2024.111798
Sanaz Katal , Benjamin York , Ali Gholamrezanezhad
While Artificial Intelligence (AI) has the potential to transform the field of diagnostic radiology, important obstacles still inhibit its integration into clinical environments. Foremost among them is the inability to integrate clinical information and prior and concurrent imaging examinations, which can lead to diagnostic errors that could irreversibly alter patient care. For AI to succeed in modern clinical practice, model training and algorithm development need to account for relevant background information that may influence the presentation of the patient in question. While AI is often remarkably accurate in distinguishing binary outcomes–hemorrhage vs. no hemorrhage; fracture vs. no fracture–the narrow scope of current training datasets prevents AI from examining the entire clinical context of the image in question. In this article, we provide an overview of the ways in which failure to account for clinical data and prior imaging can adversely affect AI interpretation of imaging studies. We then showcase how emerging techniques such as multimodal fusion and combined neural networks can take advantage of both clinical and imaging data, as well as how development strategies like domain adaptation can ensure greater generalizability of AI algorithms across diverse and dynamic clinical environments.
{"title":"AI in radiology: From promise to practice − A guide to effective integration","authors":"Sanaz Katal , Benjamin York , Ali Gholamrezanezhad","doi":"10.1016/j.ejrad.2024.111798","DOIUrl":"10.1016/j.ejrad.2024.111798","url":null,"abstract":"<div><div>While Artificial Intelligence (AI) has the potential to transform the field of diagnostic radiology, important obstacles still inhibit its integration into clinical environments. Foremost among them is the inability to integrate clinical information and prior and concurrent imaging examinations, which can lead to diagnostic errors that could irreversibly alter patient care. For AI to succeed in modern clinical practice, model training and algorithm development need to account for relevant background information that may influence the presentation of the patient in question. While AI is often remarkably accurate in distinguishing binary outcomes–hemorrhage vs. no hemorrhage; fracture vs. no fracture–the narrow scope of current training datasets prevents AI from examining the entire clinical context of the image in question. In this article, we provide an overview of the ways in which failure to account for clinical data and prior imaging can adversely affect AI interpretation of imaging studies. We then showcase how emerging techniques such as multimodal fusion and combined neural networks can take advantage of both clinical and imaging data, as well as how development strategies like domain adaptation can ensure greater generalizability of AI algorithms across diverse and dynamic clinical environments.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"181 ","pages":"Article 111798"},"PeriodicalIF":3.2,"publicationDate":"2024-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142537854","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}
Pub Date : 2024-10-19DOI: 10.1016/j.ejrad.2024.111793
Dian Zhang , Ya-nan Li , Cheng-long Li , Wan-liang Guo
Purpose
To develop a predictive model combining clinical, radiomic, and deep learning features based on X-ray and MRI to identify risk factors for early femoral head deformity in Legg-Calvé-Perthes disease (LCPD).
Methods
This study involved 152 patients diagnosed with early unilateral LCPD across two centers between January 2013 and December 2023, and included an independent external validation set to assess generalizability. Four machine learning methods, including logistic regression (LR), random forest (RF), support vector machine (SVM), and extreme gradient boosting (XGBoost), were employed to develop radiomics deep learning signatures. The clinical-radiomics model (Clinic + Rad), clinical-deep learning model (Clinic + DL), and clinical-radiomics-deep learning model (Clinic + Rad + DL) were developed by integrating radiomics deep learning signatures with clinical variables. The best model, integrated into a nomogram for clinical application, was evaluated using the area under the receiver operating characteristic curve (AUC).
Results
Among the four machine learning methods, XGBoost demonstrated superior performance in our patient dataset: radiomic (Rad) model (AUC, 0.786) and deep learning (DL) model (AUC, 0.803). Clinical variables such as age at onset and JIC classification were associated with early femoral head deformity (p < 0.05). The combined model incorporating clinical, radiomic, and deep learning signatures demonstrated better predictive ability (AUC, 0.853). The nomogram can assist clinicians in effectively assessing the risk of early femoral head deformity.
Conclusion
The Clinic + Rad + DL integrated model may be beneficial for prognostic assessment of early LCPD femoral head deformity, which is crucial for tailoring personalized treatment strategies for individual patients.
{"title":"Multimodal radiomics and deep learning models for predicting early femoral head deformity in LCPD","authors":"Dian Zhang , Ya-nan Li , Cheng-long Li , Wan-liang Guo","doi":"10.1016/j.ejrad.2024.111793","DOIUrl":"10.1016/j.ejrad.2024.111793","url":null,"abstract":"<div><h3>Purpose</h3><div>To develop a predictive model combining clinical, radiomic, and deep learning features based on X-ray and MRI to identify risk factors for early femoral head deformity in Legg-Calvé-Perthes disease (LCPD).</div></div><div><h3>Methods</h3><div>This study involved 152 patients diagnosed with early unilateral LCPD across two centers between January 2013 and December 2023, and included an independent external validation set to assess generalizability. Four machine learning methods, including logistic regression (LR), random forest (RF), support vector machine (SVM), and extreme gradient boosting (XGBoost), were employed to develop radiomics deep learning signatures. The clinical-radiomics model (Clinic + Rad), clinical-deep learning model (Clinic + DL), and clinical-radiomics-deep learning model (Clinic + Rad + DL) were developed by integrating radiomics deep learning signatures with clinical variables. The best model, integrated into a nomogram for clinical application, was evaluated using the area under the receiver operating characteristic curve (AUC).</div></div><div><h3>Results</h3><div>Among the four machine learning methods, XGBoost demonstrated superior performance in our patient dataset: radiomic (Rad) model (AUC, 0.786) and deep learning (DL) model (AUC, 0.803). Clinical variables such as age at onset and JIC classification were associated with early femoral head deformity (p < 0.05). The combined model incorporating clinical, radiomic, and deep learning signatures demonstrated better predictive ability (AUC, 0.853). The nomogram can assist clinicians in effectively assessing the risk of early femoral head deformity.</div></div><div><h3>Conclusion</h3><div>The Clinic + Rad + DL integrated model may be beneficial for prognostic assessment of early LCPD femoral head deformity, which is crucial for tailoring personalized treatment strategies for individual patients.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"181 ","pages":"Article 111793"},"PeriodicalIF":3.2,"publicationDate":"2024-10-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142497615","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}
Pub Date : 2024-10-19DOI: 10.1016/j.ejrad.2024.111797
Milan Vecsey-Nagy , Giuseppe Tremamunno , U. Joseph Schoepf , Chiara Gnasso , Emese Zsarnóczay , Nicola Fink , Dmitrij Kravchenko , Muhammad Taha Hagar , Moritz C. Halfmann , Zsófia Jokkel , Jim O’Doherty , Bálint Szilveszter , Pál Maurovich-Horvat , Pal Spruill Suranyi , Akos Varga-Szemes , Tilman Emrich
Purpose
To evaluate the feasibility of CT angiography-derived fractional flow reserve (CT-FFR) calculations on ultrahigh-resolution (UHR) photon-counting detector (PCD)-CT series and to intra-individually compare the results with energy-integrating (EID)-CT measurements.
Method
Prospective patients with calcified plaques detected on EID-CT between April 1st, 2023 and January 31st, 2024 were recruited for a UHR CCTA on PCD-CT within 30 days. PCD-CT was performed using the same or a lower CT dose index and an equivalent volume of contrast media. An on-site machine learning algorithm was used to obtain CT-FFR values on a per-vessel and per-patient basis. For all analyses, CT-FFR values ≤ 0.80 were deemed to be hemodynamically significant.
Results
A total of 34 patients (age: 67.3 ± 6.6 years, 7 women [20.6 %]) were included. Excellent inter-scanner agreement was noted for CT-FFR values in the per-vessel (ICC: 0.93 [0.90–0.95]) and per-patient (ICC: 0.94 [0.88–0.97]) analysis. PCD-CT-derived CT-FFR values proved to be higher compared to EID-CT values on both vessel (0.58 ± 0.23 vs. 0.55 ± 0.23, p < 0.001) and patient levels (0.73 ± 0.23 vs. 0.70 ± 0.22, p < 0.001). Two patients (5.9 %) with hemodynamically significant lesions on EID-CT were reclassified as non-significant on PCD-CT. All remaining participants were classified into the same category with both scanner systems.
Conclusions
While UHR CT-FFR values demonstrate excellent agreement with EID-CT measurements, PCD-CT produces higher CT-FFR values that could contribute to a reclassification of hemodynamic significance.
{"title":"Coronary CT angiography-based FFR with ultrahigh-resolution photon-counting detector CT: Intra-individual comparison to energy-integrating detector CT","authors":"Milan Vecsey-Nagy , Giuseppe Tremamunno , U. Joseph Schoepf , Chiara Gnasso , Emese Zsarnóczay , Nicola Fink , Dmitrij Kravchenko , Muhammad Taha Hagar , Moritz C. Halfmann , Zsófia Jokkel , Jim O’Doherty , Bálint Szilveszter , Pál Maurovich-Horvat , Pal Spruill Suranyi , Akos Varga-Szemes , Tilman Emrich","doi":"10.1016/j.ejrad.2024.111797","DOIUrl":"10.1016/j.ejrad.2024.111797","url":null,"abstract":"<div><h3>Purpose</h3><div>To evaluate the feasibility of CT angiography-derived fractional flow reserve (CT-FFR) calculations on ultrahigh-resolution (UHR) photon-counting detector (PCD)-CT series and to intra-individually compare the results with energy-integrating (EID)-CT measurements.</div></div><div><h3>Method</h3><div>Prospective patients with calcified plaques detected on EID-CT between April 1st, 2023 and January 31st, 2024 were recruited for a UHR CCTA on PCD-CT within 30 days. PCD-CT was performed using the same or a lower CT dose index and an equivalent volume of contrast media. An on-site machine learning algorithm was used to obtain CT-FFR values on a per-vessel and per-patient basis. For all analyses, CT-FFR values ≤ 0.80 were deemed to be hemodynamically significant.</div></div><div><h3>Results</h3><div>A total of 34 patients (age: 67.3 ± 6.6 years, 7 women [20.6 %]) were included. Excellent inter-scanner agreement was noted for CT-FFR values in the per-vessel (ICC: 0.93 [0.90–0.95]) and per-patient (ICC: 0.94 [0.88–0.97]) analysis. PCD-CT-derived CT-FFR values proved to be higher compared to EID-CT values on both vessel (0.58 ± 0.23 vs. 0.55 ± 0.23, p < 0.001) and patient levels (0.73 ± 0.23 vs. 0.70 ± 0.22, p < 0.001). Two patients (5.9 %) with hemodynamically significant lesions on EID-CT were reclassified as non-significant on PCD-CT. All remaining participants were classified into the same category with both scanner systems.</div></div><div><h3>Conclusions</h3><div>While UHR CT-FFR values demonstrate excellent agreement with EID-CT measurements, PCD-CT produces higher CT-FFR values that could contribute to a reclassification of hemodynamic significance.</div></div>","PeriodicalId":12063,"journal":{"name":"European Journal of Radiology","volume":"181 ","pages":"Article 111797"},"PeriodicalIF":3.2,"publicationDate":"2024-10-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142497612","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}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号