Radiologia Medica最新文献

{"title":"Readability versus accuracy in LLM-transformed radiology reports: stakeholder preferences across reading grade levels.","authors":"Hong-Seon Lee, Sungjun Kim, Songsoo Kim, Jeongrok Seo, Won Hwa Kim, Jaeil Kim, Kyunghwa Han, Shin Hye Hwang, Young Han Lee","doi":"10.1007/s11547-025-02098-5","DOIUrl":"https://doi.org/10.1007/s11547-025-02098-5","url":null,"abstract":"<p><strong>Purpose: </strong>To examine how reading grade levels affect stakeholder preferences based on a trade-off between accuracy and readability.</p><p><strong>Material and methods: </strong>A retrospective study of 500 radiology reports from academic and community hospitals across five imaging modalities was conducted. Reports were transformed into 11 reading grade levels (7-17) using Gemini. Accuracy, readability, and preference were rated on a 5-point scale by radiologists, physicians, and laypersons. Errors (generalizations, omissions, hallucinations) and potential changes in patient management (PCPM) were identified. Ordinal logistic regression analyzed preference predictors, and weighted kappa measured interobserver reliability.</p><p><strong>Results: </strong>Preferences varied across reading grade levels depending on stakeholder group, modality, and clinical setting. Overall, preferences peaked at grade 16, but declined at grade 17, particularly among laypersons. Lower reading grades improved readability but increased errors, while higher grades improved accuracy but reduced readability. In multivariable analysis, accuracy was the strongest predictor of preference for all groups (OR: 30.29, 33.05, and 2.16; p <0 .001), followed by readability (OR: 2.73, 1.70, 2.01; p <0.001).</p><p><strong>Conclusion: </strong>Higher-grade levels were generally preferred due to better accuracy, with a range of 12-17. Further increasing grade levels reduced readability sharply, limiting preference. These findings highlight the limitations of unsupervised LLM transformations and suggest the need for hybrid approaches that maintain original reports while incorporating explanatory content to balance accuracy and readability.</p>","PeriodicalId":20817,"journal":{"name":"Radiologia Medica","volume":" ","pages":""},"PeriodicalIF":4.8,"publicationDate":"2025-09-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145192611","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Target node selection and pathologic correlation in post-vascular CEUS for axillary staging in early-stage breast cancer.","authors":"Deniz Esin Tekcan Sanli, Ahmet Necati Sanli","doi":"10.1007/s11547-025-02108-6","DOIUrl":"https://doi.org/10.1007/s11547-025-02108-6","url":null,"abstract":"","PeriodicalId":20817,"journal":{"name":"Radiologia Medica","volume":" ","pages":""},"PeriodicalIF":4.8,"publicationDate":"2025-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145182276","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Reply to the letter to the editor \"preoperative imaging in breast cancer staging: can CEM stand alone?\"","authors":"Giulia Bicchierai, Francesco Amato, Chiara Bellini, Jacopo Nori","doi":"10.1007/s11547-025-02095-8","DOIUrl":"https://doi.org/10.1007/s11547-025-02095-8","url":null,"abstract":"","PeriodicalId":20817,"journal":{"name":"Radiologia Medica","volume":" ","pages":""},"PeriodicalIF":4.8,"publicationDate":"2025-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145150658","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Ultra-low-field MRI: a David versus Goliath challenge in modern imaging.","authors":"Cesare Gagliardo, Paola Feraco, Eleonora Contrino, Costanza D'Angelo, Laura Geraci, Giuseppe Salvaggio, Andrea Gagliardo, Ludovico La Grutta, Massimo Midiri, Maurizio Marrale","doi":"10.1007/s11547-025-02091-y","DOIUrl":"https://doi.org/10.1007/s11547-025-02091-y","url":null,"abstract":"<p><p>Ultra-low-field magnetic resonance imaging (ULF-MRI), operating below 0.2 Tesla, is gaining renewed interest as a re-emerging diagnostic modality in a field dominated by high- and ultra-high-field systems. Recent advances in magnet design, RF coils, pulse sequences, and AI-based reconstruction have significantly enhanced image quality, mitigating traditional limitations such as low signal- and contrast-to-noise ratio and reduced spatial resolution. ULF-MRI offers distinct advantages: reduced susceptibility artifacts, safer imaging in patients with metallic implants, low power consumption, and true portability for point-of-care use. This narrative review synthesizes the physical foundations, technological advances, and emerging clinical applications of ULF-MRI. A focused literature search across PubMed, Scopus, IEEE Xplore, and Google Scholar was conducted up to August 11, 2025, using combined keywords targeting hardware, software, and clinical domains. Inclusion emphasized scientific rigor and thematic relevance. A comparative analysis with other imaging modalities highlights the specific niche ULF-MRI occupies within the broader diagnostic landscape. Future directions and challenges for clinical translation are explored. In a world increasingly polarized between the push for ultra-high-field excellence and the need for accessible imaging, ULF-MRI embodies a modern \"David versus Goliath\" theme, offering a sustainable, democratizing force capable of expanding MRI access to anyone, anywhere.</p>","PeriodicalId":20817,"journal":{"name":"Radiologia Medica","volume":" ","pages":""},"PeriodicalIF":4.8,"publicationDate":"2025-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145150635","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Diagnostic performance based on MRI for preoperative of vessels encapsulating tumor clusters in hepatocellular carcinoma: a systematic review and meta-analysis.","authors":"Miaomiao Wang, Yinzhong Wang, Liang Cao, Qian Wang, Ya Shen, Xiaoxue Tian, Junqiang Lei","doi":"10.1007/s11547-025-02100-0","DOIUrl":"https://doi.org/10.1007/s11547-025-02100-0","url":null,"abstract":"<p><strong>Objective: </strong>Vessels encapsulating tumor clusters (VETC) pattern is a unique pattern of vascular invasion that has been shown to be a poor prognostic factor for hepatocellular carcinoma (HCC). The purpose of this review and meta-analysis was to explore diagnostic performance between non-radiomics and radiomics model based on MRI for preoperative of vessels encapsulating tumor clusters (VETC) in hepatocellular carcinoma (HCC).</p><p><strong>Methods: </strong>All included articles were obtained from PubMed, Embase, Web of Science and Cochrane library as of September 30, 2024. The QUADAS-2 tool was used to assess methodological quality of eligible studies. The pooled data was using a mixed effects model within a 95% confidence interval (CI). Diagnostic performance was represented by summary receiver-operating characteristic curves and the area under the curve (AUC).</p><p><strong>Results: </strong>A total of 14 studies (10 non-radiomics and 6 radiomics) with 2961 HCC patients were included in this study. The pooled sensitivity and specificity of non radiomics model were 0.80 (95%CI:0.76-0.83) and 0.74(95%CI:0.69-0.78), with AUC of 0.84 (95%CI:0.80-0.87); whereas that of radiomics model was 0.88 (95%CI:0.83- 0.91) and 0.86 (95%CI:0.81-0.90) with AUC of 0.93 (95%CI:0.91-0.95).</p><p><strong>Conclusions: </strong>Radiomics model performed better than non-radiomics model based on MRI in preoperative prediction of VETC-positive HCC, but there was heterogeneity between studies, which needs to be interpreted with caution.</p>","PeriodicalId":20817,"journal":{"name":"Radiologia Medica","volume":" ","pages":""},"PeriodicalIF":4.8,"publicationDate":"2025-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145131760","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"In-context learning enables large language models to achieve human-level performance in spinal instability neoplastic score classification from synthetic CT and MRI reports.","authors":"Maximilian F Russe, Marco Reisert, Anna Fink, Marc Hohenhaus, Julia M Nakagawa, Caroline Wilpert, Carl P Simon, Elmar Kotter, Horst Urbach, Alexander Rau","doi":"10.1007/s11547-025-02096-7","DOIUrl":"https://doi.org/10.1007/s11547-025-02096-7","url":null,"abstract":"<p><strong>Purpose: </strong>To assess the performance of state-of-the-art large language models in classifying vertebral metastasis stability using the Spinal Instability Neoplastic Score (SINS) compared to human experts, and to evaluate the impact of task-specific refinement including in-context learning on their performance.</p><p><strong>Material and methods: </strong>This retrospective study analyzed 100 synthetic CT and MRI reports encompassing a broad range of SINS scores. Four human experts (two radiologists and two neurosurgeons) and four large language models (Mistral, Claude, GPT-4 turbo, and GPT-4o) evaluated the reports. Large language models were tested in both generic form and with task-specific refinement. Performance was assessed based on correct SINS category assignment and attributed SINS points.</p><p><strong>Results: </strong>Human experts demonstrated high median performance in SINS classification (98.5% correct) and points calculation (92% correct), with a median point offset of 0 [0-0]. Generic large language models performed poorly with 26-63% correct category and 4-15% correct SINS points allocation. In-context learning significantly improved chatbot performance to near-human levels (96-98/100 correct for classification, 86-95/100 for scoring, no significant difference to human experts). Refined large language models performed 71-85% better in SINS points allocation.</p><p><strong>Conclusion: </strong>In-context learning enables state-of-the-art large language models to perform at near-human expert levels in SINS classification, offering potential for automating vertebral metastasis stability assessment. The poor performance of generic large language models highlights the importance of task-specific refinement in medical applications of artificial intelligence.</p>","PeriodicalId":20817,"journal":{"name":"Radiologia Medica","volume":" ","pages":""},"PeriodicalIF":4.8,"publicationDate":"2025-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145131753","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Response to the letter: \"deep learning-driven knee MRI acceleration: evidence, enhancements, and the path to universal adoption\".","authors":"Giovanni Foti, Flavio Spoto, Alessandro Spezia, Simone Caia","doi":"10.1007/s11547-025-02101-z","DOIUrl":"https://doi.org/10.1007/s11547-025-02101-z","url":null,"abstract":"","PeriodicalId":20817,"journal":{"name":"Radiologia Medica","volume":" ","pages":""},"PeriodicalIF":4.8,"publicationDate":"2025-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145126139","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Radiomics integrated with machine and deep learning analysis of T2-weighted and arterial-phase T1-weighted Magnetic Resonance Imaging for non-invasive detection of metastatic axillary lymph nodes in breast cancer.","authors":"Roberta Fusco, Vincenza Granata, Mauro Mattace Raso, Igino Simonetti, Paolo Vallone, Davide Pupo, Filippo Tovecci, Maria Assunta Daniela Iasevoli, Francesca Maio, Paola Gargiulo, Giuditta Giannotti, Paolo Pariante, Saverio Simonelli, Gerardo Ferrara, Claudio Siani, Raimondo Di Giacomo, Sergio Venanzio Setola, Antonella Petrillo","doi":"10.1007/s11547-025-02090-z","DOIUrl":"https://doi.org/10.1007/s11547-025-02090-z","url":null,"abstract":"<p><strong>Purpose: </strong>To compare the diagnostic performance of radiomic features extracted from T2-weighted and arterial-phase T1-weighted MRI sequences using univariate, machine and deep learning analysis and to assess their effectiveness in predicting axillary lymph node (ALN) metastasis in breast cancer patients.</p><p><strong>Methods: </strong>We retrospectively analyzed MRI data from 100 breast cancer patients, comprising 52 metastatic and 103 non-metastatic lymph nodes. Radiomic features were extracted from T2-weighted and subtracted arterial-phase T1-weighted images. Feature normalization and selection were performed. Various machine learning classifiers, including logistic regression, gradient boosting, random forest, and neural networks, were trained and evaluated. Diagnostic performance was assessed using metrics such as area under the curve (AUC), sensitivity, specificity, and accuracy.</p><p><strong>Results: </strong>T2-weighted imaging provided strong performance in multivariate modeling, with the neural network achieving the highest AUC (0.978) and accuracy (91.1%), showing statistically significant differences over models. The stepwise logistic regression model also showed competitive results (AUC = 0.796; accuracy = 73.3%). In contrast, arterial-phase T1-weighted imaging features performed better when analyzed individually, with the best univariate AUC reaching 0.787. When multivariate modeling was applied to arterial-phase features, the best-performing logistic regression model achieved an AUC of 0.853 and accuracy of 77.8%.</p><p><strong>Conclusion: </strong>Radiomic analysis of T2-weighted MRI, particularly through deep learning models like neural networks, demonstrated the highest overall diagnostic performance for predicting metastatic ALNs. In contrast, arterial-phase T1-weighted features showed better results in univariate analysis. These findings support the integration of radiomic features, especially from T2-weighted sequences, into multivariate models to enhance noninvasive preoperative assessment.</p>","PeriodicalId":20817,"journal":{"name":"Radiologia Medica","volume":" ","pages":""},"PeriodicalIF":4.8,"publicationDate":"2025-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145126070","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Preoperative Imaging in breast cancer staging: Can CEM stand alone?","authors":"Deniz Esin Tekcan Sanli, Ahmet Necati Sanli","doi":"10.1007/s11547-025-02093-w","DOIUrl":"https://doi.org/10.1007/s11547-025-02093-w","url":null,"abstract":"","PeriodicalId":20817,"journal":{"name":"Radiologia Medica","volume":" ","pages":""},"PeriodicalIF":4.8,"publicationDate":"2025-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145126101","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep learning-driven knee MRI acceleration: evidence, enhancements, and the path to universal adoption.","authors":"Yanxia Jia, Shan Tao, Chengqiang Jin","doi":"10.1007/s11547-025-02092-x","DOIUrl":"https://doi.org/10.1007/s11547-025-02092-x","url":null,"abstract":"","PeriodicalId":20817,"journal":{"name":"Radiologia Medica","volume":" ","pages":""},"PeriodicalIF":4.8,"publicationDate":"2025-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145126067","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}