Academic Radiology最新文献

{"title":"Multiparametric MRI-based Machine Learning Radiomics for Predicting Treatment Response to Transarterial Chemoembolization Combined with Targeted and Immunotherapy in Unresectable Hepatocellular Carcinoma: A Multicenter Study","authors":"Wendi Kang ,&nbsp;Peiyun Tang ,&nbsp;Yingen Luo ,&nbsp;Qicai Lian ,&nbsp;Xuan Zhou ,&nbsp;Jinrui Ren ,&nbsp;Tianhao Cong ,&nbsp;Lei Miao ,&nbsp;Hang Li ,&nbsp;Xiaoyu Huang ,&nbsp;Aixin Ou ,&nbsp;Hao Li ,&nbsp;Zhentao Yan ,&nbsp;Yingjie Di ,&nbsp;Xiao Li ,&nbsp;Feng Ye ,&nbsp;Xiaoli Zhu ,&nbsp;Zhengqiang Yang","doi":"10.1016/j.acra.2024.10.038","DOIUrl":"10.1016/j.acra.2024.10.038","url":null,"abstract":"<div><h3>Rationale and Objectives</h3><div>To develop and validate multiple machine learning predictive models incorporating clinical features and pretreatment multiparametric magnetic resonance imaging (MRI) radiomic features for predicting treatment response to transarterial chemoembolization combined with molecular targeted therapy plus immunotherapy in unresectable hepatocellular carcinoma (HCC).</div></div><div><h3>Materials and methods</h3><div>This retrospective study involved 276 patients with unresectable HCC who received combination therapy from 4 medical centers. Patients were divided into one training cohort and two independent external validation cohorts. 16 radiomic features from six multiparametric MRI sequences and 2 clinical features were used to build six machine learning models. The models were evaluated using the area under the curve (AUC), decision curve analysis, and incremental predictive value.</div></div><div><h3>Results</h3><div>Alpha-fetoprotein and neutrophil-to-lymphocyte ratio are clinical independent predictors of treatment response. In the training cohort and two external validation cohorts, the AUCs and 95% confidence intervals for predicting treatment response were respectively 0.782 (0.698-0.857) 0.695 (0.566–0.823), and 0.679 (0.542–0.810) for the clinical model; 0.942 (0.903–0.974), 0.869 (0.761–0.949), and 0.868 (0.769–0.942) for the radiomics model; and 0.956 (0.920–0.984), 0.895 (0.810–0.967), and 0.892 (0.804–0.957) for the combined clinical-radiomics model. In the three cohorts, the incremental predictive value of the radiomics model over the clinical model was 49.2% (P &lt; 0.001), 28.8% (P &lt; 0.001), and 31.5% (P &lt; 0.001).</div></div><div><h3>Conclusion</h3><div>The combined clinical-radiomics model may provide a reliable and non-invasive tool to predict individual treatment responses and guide and improve clinical decision-making in combination therapy of HCC patients.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"32 4","pages":"Pages 2013-2026"},"PeriodicalIF":3.8,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142752234","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Feasibility of Sub-milliSievert Low-dose Computed Tomography with Deep Learning Image Reconstruction in Evaluating Pulmonary Subsolid Nodules: A Prospective Intra-individual Comparison Study","authors":"Huiyuan Zhu ,&nbsp;Zike Huang ,&nbsp;Qunhui Chen ,&nbsp;Weiling Ma ,&nbsp;Jiahui Yu ,&nbsp;Shiqing Wang ,&nbsp;Guangyu Tao ,&nbsp;Jun Xing ,&nbsp;Haixin Jiang ,&nbsp;Xiwen Sun ,&nbsp;Jing Liu ,&nbsp;Hong Yu ,&nbsp;Lin Zhu","doi":"10.1016/j.acra.2024.11.042","DOIUrl":"10.1016/j.acra.2024.11.042","url":null,"abstract":"<div><h3>Rationale and Objectives</h3><div>To comprehensively assess the feasibility of low-dose computed tomography (LDCT) using deep learning image reconstruction (DLIR) for evaluating pulmonary subsolid nodules, which are challenging due to their susceptibility to noise.</div></div><div><h3>Materials and Methods</h3><div>Patients undergoing both standard-dose CT (SDCT) and LDCT between March and June 2023 were prospectively enrolled. LDCT images were reconstructed with high-strength DLIR (DLIR-H), medium-strength DLIR (DLIR-M), adaptive statistical iterative reconstruction-V level 50% (ASIR-V-50%), and filtered back projection (FBP); SDCT with FBP as the reference standard. Objective assessment, including image noise, contrast-to-noise ratio (CNR), and signal-to-noise ratio (SNR), and subjective assessment using five-point scales by five radiologists were performed. Detection and false-positive rate of subsolid nodules, and morphologic features of nodules were recorded.</div></div><div><h3>Results</h3><div>102 patients (mean age, 57.0 ± 12.3 years) with 358 subsolid nodules in SDCT were enrolled. The mean effective dose of SDCT and LDCT were 5.37 ± 0.80<!--> <!-->mSv and 0.86 ± 0.14<!--> <!-->mSv, respectively (<em>P</em> &lt; 0.001). DLIR-H showed the lowest noise, highest CNRs, SNRs, and subjective scores among LDCT groups (all <em>P</em> &lt; 0.001), almost approaching comparability with SDCT. The detection rates for DLIR-H, DLIR-M, ASIR-V-50%, and FBP were 76.5%, 76.3%, 83.8%, and 72.1%, respectively (<em>P</em> &lt; 0.001), with false-positive rate of 2.5%, 2.2%, 8.3%, and 1.1%, respectively (<em>P</em> &lt; 0.001). DLIR-H showed the highest detection rates for morphologic features (79.4%–95.2%) compared to DLIR-M (74.6%–88.9%), ASIR-V-50% (72.0%–88.4%), and FBP (66.1%–84.1%) (all <em>P</em> ≤ 0.001).</div></div><div><h3>Conclusion</h3><div>Sub-milliSievert LDCT with DLIR-H offers substantial dose reduction without compromising image quality. It is promising for evaluating subsolid nodules with a high detection rate and better identification of morphologic features.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"32 4","pages":"Pages 2309-2319"},"PeriodicalIF":3.8,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142824692","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Early Lung Adenocarcinoma Manifesting as Irregular Subsolid Nodules: Clinical and CT Characteristics","authors":"Pei-ling Zou MD ,&nbsp;Chao-hao Ma MD ,&nbsp;Xian Li PhD ,&nbsp;Tian-you Luo PhD ,&nbsp;Fa-jin Lv PhD ,&nbsp;Qi Li PhD","doi":"10.1016/j.acra.2024.12.010","DOIUrl":"10.1016/j.acra.2024.12.010","url":null,"abstract":"<div><h3>Rationale and Objectives</h3><div>To explore the clinical and computed tomography (CT) characteristics of early-stage lung adenocarcinoma (LADC) that presents with an irregular shape.</div></div><div><h3>Materials and Methods</h3><div>The CT data of 575 patients with stage IA LADC and 295 with persistent inflammatory lesion (PIL) manifesting as subsolid nodules (SSNs) were analyzed retrospectively. Among these patients, we selected 233 patients with LADC and 140 patients with PIL, who showed irregular SSNs, hereinafter referred to as irregular LADC (I-LADC) and irregular PIL (I-PIL), respectively. The incidence rates, clinical characteristics, and CT features of I-LADC and I-PIL were compared. Additionally, binary logistic regression analysis was performed to determine the independent factors for diagnosing I-LADC.</div></div><div><h3>Results</h3><div>The incidence rates of I-LADC and I-PIL were 40.5% (233/575) and 47.5% (140/295), respectively, with no statistically significant difference observed between the two groups (<em>P</em> &gt; 0.05). Univariate analysis revealed significant differences in three clinical characteristics and 13 radiological features between I-LADC and I-PIL (all <em>P</em> &lt; 0.05). Binary logistic regression indicated that the alignment of the long axis of SSN with the bronchial vascular bundle, a well-defined boundary of ground-glass opacity, lobulation, arc concave sign, and absence of knife-like change were the independent predictors of I-LADC, yielding an area under the curve and accuracy of 0.979% and 93.5%, respectively.</div></div><div><h3>Conclusion</h3><div>Early LADC presenting as SSNs is associated with a high incidence of irregular shape. I-LADC and I-PIL exhibited different clinical and imaging characteristics. A good understanding of these differences may be helpful for the accurate diagnosis of I-LADC.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"32 4","pages":"Pages 2320-2329"},"PeriodicalIF":3.8,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142898924","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Incidental Hypermetabolic Breast Lesions on 18F-FDG PET-CT: Clinical Significance, Diagnostic Strategies, and Future Directions","authors":"Derek L. Nguyen MD","doi":"10.1016/j.acra.2025.03.003","DOIUrl":"10.1016/j.acra.2025.03.003","url":null,"abstract":"","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"32 4","pages":"Pages 1816-1817"},"PeriodicalIF":3.8,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143574578","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The Value of Second-look Ultrasound and Mammography for Assessment and Biopsy of MRI-detected Breast Lesions","authors":"Stephanie Tina Sauer MD ,&nbsp;Julius Geerling ,&nbsp;Sara Aniki Christner MD ,&nbsp;Tanja Schlaiß MD ,&nbsp;Matthias Kiesel MD ,&nbsp;Anne Cathrine Scherer-Quenzer MD ,&nbsp;Lukas Müller MD ,&nbsp;Julius Frederik Heidenreich MD ,&nbsp;Thorsten Alexander Bley MD ,&nbsp;Jan-Peter Grunz MD","doi":"10.1016/j.acra.2024.10.037","DOIUrl":"10.1016/j.acra.2024.10.037","url":null,"abstract":"<div><h3>Rationale and Objectives</h3><div>Suspicious lesions detected in multiparametric breast MRI can be further analyzed with second-look ultrasound (SLUS) and/or mammography. This study aims to assess the value of second-look imaging in selecting the appropriate biopsy method for different lesion characteristics.</div></div><div><h3>Materials and Methods</h3><div>Between January 2021 and December 2023, 212 women underwent contrast-enhanced multiparametric breast MRI at 3 Tesla. A total of 241 suspicious lesions (108 malignancies, 44.8%) were further assessed with SLUS and second-look mammography. Subsequent image-guided biopsy of each lesion was performed using the most suitable modality. Size-dependent lesion detection rates in SLUS and mammography were compared by means of the McNemar test.</div></div><div><h3>Results</h3><div>Lesions referred to MRI-guided biopsy were predominantly ≤ 10 mm in size (52.8%). SLUS allowed for higher detection rates than mammography in mass lesions (55.6% [95% confidence interval 46.4–64.4%] versus 16.7% [10.6–24.3%]; p &lt; 0.001) with a particularly high sensitivity for malignant mass lesions &gt; 10 mm (88.5% [69.9–97.6%]). In contrast, the detection rate for malignant non-mass lesions was lower in SLUS than in second-look mammography (22.0% [11.5–36.0%] versus 38.0% [24.7–52.8%]; p &lt; 0.001). The malignancy rates in ultrasound-, mammography-, and MRI-guided biopsies were 53.7%, 55.2%, and 35.0%, respectively.</div></div><div><h3>Conclusion</h3><div>SLUS is an excellent tool for further assessment and biopsy of suspicious mass lesions &gt; 10 mm without associated calcifications. In contrast, supplemental ultrasound is of limited value in the evaluation and biopsy guidance of suspicious non-mass lesions compared to second-look mammography.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"32 4","pages":"Pages 1818-1826"},"PeriodicalIF":3.8,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142606712","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The Value of Machine Learning-based Radiomics Model Characterized by PET Imaging with 68Ga-FAPI in Assessing Microvascular Invasion of Hepatocellular Carcinoma","authors":"Rongqin Fan ,&nbsp;Xueqin Long ,&nbsp;Xiaoliang Chen ,&nbsp;Yanmei Wang ,&nbsp;Demei Chen ,&nbsp;Rui Zhou","doi":"10.1016/j.acra.2024.11.034","DOIUrl":"10.1016/j.acra.2024.11.034","url":null,"abstract":"<div><h3>Rationale and Objectives</h3><div>This study aimed to develop a radiomics model characterized by ⁶⁸Ga-fibroblast activation protein inhibitors (FAPI) positron emission tomography (PET) imaging to predict microvascular invasion (MVI) of hepatocellular carcinoma (HCC). This study also investigated the impact of varying thresholds for maximum standardized uptake value (SUV_max) in semi-automatic delineation methods on the predictions of the model.</div></div><div><h3>Methods</h3><div>This retrospective study included 84 HCC patients who underwent ⁶⁸Ga-FAPI PET and their MVI results were confirmed by histopathological examination. Volumes of interest (VOIs) for lesions were semi-automatically delineated with four thresholds of 30%, 40%, 50%, and 60% for SUV_max. Extracted shape features, first-, second- and higher-order features. Eight PET radiomics models for predicting MVI were constructed and tested.</div></div><div><h3>Results</h3><div>In the testing set, the logistic regression (LR) model achieved the highest AUC values for three groups of 30%, 50%, and 60%, with values of 0.785, 0.896, and 0.859, respectively, while the random forest (RF) model in 40% group obtained the highest AUC value of 0.815. The LR model in 50% group and the extreme gradient boosting (XGBoost) model in 60% group achieved the highest accuracy, each at 87.5%. The highest sensitivity was observed in the support vector machine (SVM) model in 30% group, at 100%.</div></div><div><h3>Conclusion</h3><div>The ⁶⁸Ga-FAPI PET radiomics model has high efficacy in predicting MVI in HCC, which is important for the development of HCC treatment plan and post-treatment evaluation. Different thresholds of SUV_max in semi-automatic delineation methods exert a degree of influence on performance of the radiomics model.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"32 4","pages":"Pages 2233-2246"},"PeriodicalIF":3.8,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142796527","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Baseline Body Composition and 3D-Extracellular Volume Fraction for Predicting Pancreatic Fistula after Distal Pancreatectomy in Pancreatic Body and/or Tail Adenocarcinoma","authors":"Wei Cai ,&nbsp;Yongjian Zhu ,&nbsp;Dengfeng Li ,&nbsp;Mancang Hu ,&nbsp;Ze Teng ,&nbsp;Rong Cong ,&nbsp;Zhaowei Chen ,&nbsp;Xujie Sun ,&nbsp;Xiaohong Ma ,&nbsp;Xinming Zhao","doi":"10.1016/j.acra.2024.10.010","DOIUrl":"10.1016/j.acra.2024.10.010","url":null,"abstract":"<div><h3>Rationale and Objectives</h3><div>Clinically relevant postoperative pancreatic fistula (CR-POPF) is a threatening complication in body and/or tail pancreatic ductal adenocarcinoma (PDAC) receiving distal pancreatectomy (DP) and is difficult to predict preoperatively. We aimed to identify the role of baseline CT-based body composition analysis and extracellular volume (ECV) map in predicting the risk of CR-POPF preoperatively.</div></div><div><h3>Materials and Methods</h3><div>A total of 329 resectable PDAC patients were enrolled and underwent multiphasic contrast-enhanced CT. Body composition indicators were calculated, and ECV maps were generated through multiphasic CT images. The differences in clinical variables and quantitative parameters between CR-POPF and non-CR-POPF patients were compared. Correlations between ECV fraction and pancreatic fibrosis stage were analyzed. Multivariate logistic regression was performed to screen the independent predictors and develop prediction models for CR-POPF. Receiver operating characteristic curve was utilized to evaluate the predictive performance.</div></div><div><h3>Results</h3><div>Among 329 patients, 19.76% (65/329) developed CR-POPF. Albumin, pancreatic texture, and intraoperative blood loss were used to build the clinical model with an AUC of 0.764. ECV fraction and total muscle ratio (TMR) were chosen to build the radiological model with an AUC of 0.872. A combined nomogram integrated with albumin, ECV fraction, and TMR could significantly improve the discrimination ability to an AUC of 0.924 (Delong test, all <em>p</em> &lt; 0.05). The ECV fraction showed high positive correlation with histological fibrosis grade (Spearman ρ = 0.81).</div></div><div><h3>Conclusion</h3><div>CT-based body composition analysis and ECV exhibited great potential for predicting CR-POPF in body and/or tail PDAC after DP. The combined nomogram could further improve the predictive performance.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"32 4","pages":"Pages 2027-2040"},"PeriodicalIF":3.8,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142631582","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Evaluating the Efficacy of Perplexity Scores in Distinguishing AI-Generated and Human-Written Abstracts","authors":"Alperen Elek ,&nbsp;Hatice Sude Yildiz ,&nbsp;Benan Akca ,&nbsp;Nisa Cem Oren ,&nbsp;Batuhan Gundogdu","doi":"10.1016/j.acra.2025.01.017","DOIUrl":"10.1016/j.acra.2025.01.017","url":null,"abstract":"<div><h3>Rationale and Objectives</h3><div>We aimed to evaluate the efficacy of perplexity scores in distinguishing between human-written and AI-generated radiology abstracts and to assess the relative performance of available AI detection tools in detecting AI-generated content.</div></div><div><h3>Methods</h3><div>Academic articles were curated from PubMed using the keywords \"neuroimaging\" and \"angiography.\" Filters included English-language, open-access articles with abstracts without subheadings, published before 2021, and within Chatbot processing word limits. The first 50 qualifying articles were selected, and their full texts were used to create AI-generated abstracts. Perplexity scores, which estimate sentence predictability, were calculated for both AI-generated and human-written abstracts. The performance of three AI tools in discriminating human-written from AI-generated abstracts was assessed.</div></div><div><h3>Results</h3><div>The selected 50 articles consist of 22 review articles (44%), 12 case or technical reports (24%), 15 research articles (30%), and one editorial (2%). The perplexity scores for human-written abstracts (median; 35.9 IQR; 25.11–51.8) were higher than those for AI-generated abstracts (median; 21.2 IQR; 16.87–28.38), (p<!--> <!-->=<!--> <!-->0.057) with an AUC<!--> <!-->=<!--> <!-->0.7794. One AI tool performed less than chance in identifying human-written from AI-generated abstracts with an accuracy of 36% (p<!--> <!-->&gt;<!--> <!-->0.05) while another tool yielded an accuracy of 95% with an AUC<!--> <!-->=<!--> <!-->0.8688.</div></div><div><h3>Conclusion</h3><div>This study underscores the potential of perplexity scores in detecting AI-generated and potentially fraudulent abstracts. However, more research is needed to further explore these findings and their implications for the use of AI in academic writing. Future studies could also investigate other metrics or methods for distinguishing between human-written and AI-generated texts.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"32 4","pages":"Pages 1785-1790"},"PeriodicalIF":3.8,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143366773","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Currently Available Radiology-Specific Reporting Guidelines","authors":"Paul Cronin MD, MS ,&nbsp;Omar Msto Hussain Nasser MD ,&nbsp;James V. Rawson MD","doi":"10.1016/j.acra.2025.01.014","DOIUrl":"10.1016/j.acra.2025.01.014","url":null,"abstract":"<div><div>The aim of this paper is to contextualize and review reporting guidelines available at the EQUATOR Network that are most relevant to radiology-specific investigations. Eight EQUATOR Network reporting guidelines for the clinical area of radiology, not including the subspecialized areas of imaging of the cardiovascular, neurologic, and oncologic diseases are reviewed and discussed. The reporting guidelines are for diagnostic and therapeutic clinical research. Why the reporting guideline was development, by whom, their aims and what they hope to achieve are discussed. A table summarizes what the reporting guideline is provided for; an acronym if present is given; a full bibliographic reference with PMID number; the reporting guideline website URL or link; the study design and section of the report that the guideline applies to; and the date that the reporting guideline was last updated.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"32 4","pages":"Pages 1798-1805"},"PeriodicalIF":3.8,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143069333","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Decision Fusion Model for Predicting Microvascular Invasion in Hepatocellular Carcinoma Based on Multi-MR Habitat Imaging and Machine-Learning Classifiers","authors":"Zhenhuan Huang ,&nbsp;Wanrong Huang ,&nbsp;Lu Jiang ,&nbsp;Yao Zheng ,&nbsp;Yifan Pan ,&nbsp;Chuan Yan ,&nbsp;Rongping Ye ,&nbsp;Shuping Weng ,&nbsp;Yueming Li MD","doi":"10.1016/j.acra.2024.10.007","DOIUrl":"10.1016/j.acra.2024.10.007","url":null,"abstract":"<div><h3>Rationale and Objectives</h3><div>Accurate prediction of microvascular invasion (MVI) in hepatocellular carcinoma (HCC) is crucial for guiding treatment. This study evaluates and compares the performance of clinicoradiologic, traditional radiomics, deep-learning radiomics, feature fusion, and decision fusion models based on multi-region MR habitat imaging using six machine-learning classifiers.</div></div><div><h3>Materials and Methods</h3><div>We retrospectively included 300 HCC patients. The intratumoral and peritumoral regions were segmented into distinct habitats, from which radiomics and deep-learning features were extracted using arterial phase MR images. To reduce feature dimensionality, we applied intra-class correlation coefficient (ICC) analysis, Pearson correlation coefficient (PCC) filtering, and recursive feature elimination (RFE). Based on the selected optimal features, prediction models were constructed using decision tree (DT), K-nearest neighbors (KNN), logistic regression (LR), random forest (RF), support vector machine (SVM), and XGBoost (XGB) classifiers. Additionally, fusion models were developed utilizing both feature fusion and decision fusion strategies. The performance of these models was validated using the area under the receiver operating characteristic curve (ROC AUC), calibration curves, and decision curve analysis.</div></div><div><h3>Results</h3><div>The decision fusion model (VOI-Peri10–1) using LR and integrating clinicoradiologic, radiomics, and deep-learning features achieved the highest AUC of 0.808 (95% confidence interval [CI]: 0.807–0.912) in the test cohort, with good calibration (Hosmer–Lemeshow test, <em>P</em> &gt; 0.050) and clinical net benefit.</div></div><div><h3>Conclusion</h3><div>The LR-based decision fusion model integrating clinicoradiologic, radiomics, and deep-learning features shows promise for preoperative prediction of MVI in HCC, aiding in patient outcome predictions and personalized treatment planning.</div></div>","PeriodicalId":50928,"journal":{"name":"Academic Radiology","volume":"32 4","pages":"Pages 1971-1980"},"PeriodicalIF":3.8,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142548784","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}