Radiology-Artificial Intelligence最新文献

{"title":"Applying Conformal Prediction to a Deep Learning Model for Intracranial Hemorrhage Detection to Improve Trustworthiness.","authors":"Cooper Gamble, Shahriar Faghani, Bradley J Erickson","doi":"10.1148/ryai.240032","DOIUrl":"https://doi.org/10.1148/ryai.240032","url":null,"abstract":"<p><p><i>\"Just Accepted\" papers have undergone full peer review and have been accepted for publication in <i>Radiology: Artificial Intelligence</i>. This article will undergo copyediting, layout, and proof review before it is published in its final version. Please note that during production of the final copyedited article, errors may be discovered which could affect the content.</i> Purpose To apply conformal prediction to a deep learning (DL) model for intracranial hemorrhage (ICH) detection and evaluate model performance in detection as well as model accuracy in identifying challenging cases. Materials and Methods This was a retrospective (November 2017 through December 2017) study of 491 noncontrast head CT volumes from the CQ500 dataset in which three senior radiologists annotated sections containing ICH. The dataset was split into definite and challenging (uncertain) subsets, where challenging images were defined as those in which there was disagreement among readers. A DL model was trained on 146 patients (mean age = 45.7, 70 females, 76 males) from the definite data (training dataset) to perform ICH localization and classification into five classes. To develop an uncertainty-aware DL model, 1,546 sections of the definite data (calibration dataset) was used for Mondrian conformal prediction (MCP). The uncertainty-aware DL model was tested on 8,401 definite and challenging sections to assess its ability to identify challenging sections. The difference in predictive performance (<i>P</i> value) and ability to identify challenging sections (accuracy) were reported. Results After the MCP procedure, the model achieved an F1 score of 0.920 for ICH classification on the test dataset. Additionally, it correctly identified 6,837 of the 6,856 total challenging sections as challenging (99.7% accuracy). It did not incorrectly label any definite sections as challenging. Conclusion The uncertainty-aware MCP-augmented DL model achieved high performance in ICH detection and high accuracy in identifying challenging sections, suggesting its usefulness in automated ICH detection and potential to increase trustworthiness of DL models in radiology. ©RSNA, 2024.</p>","PeriodicalId":29787,"journal":{"name":"Radiology-Artificial Intelligence","volume":" ","pages":"e240032"},"PeriodicalIF":8.1,"publicationDate":"2024-11-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142732987","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Achieving More with Less: Combining Strong and Weak Labels for Intracranial Hemorrhage Detection.","authors":"Tugba Akinci D'Antonoli, Jeffrey D Rudie","doi":"10.1148/ryai.240670","DOIUrl":"10.1148/ryai.240670","url":null,"abstract":"","PeriodicalId":29787,"journal":{"name":"Radiology-Artificial Intelligence","volume":"6 6","pages":"e240670"},"PeriodicalIF":8.1,"publicationDate":"2024-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11605141/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142584303","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":"Addressing the Generalizability of AI in Radiology Using a Novel Data Augmentation Framework with Synthetic Patient Image Data: Proof-of-Concept and External Validation for Classification Tasks in Multiple Sclerosis.","authors":"Gianluca Brugnara, Chandrakanth Jayachandran Preetha, Katerina Deike, Robert Haase, Thomas Pinetz, Martha Foltyn-Dumitru, Mustafa A Mahmutoglu, Brigitte Wildemann, Ricarda Diem, Wolfgang Wick, Alexander Radbruch, Martin Bendszus, Hagen Meredig, Aditya Rastogi, Philipp Vollmuth","doi":"10.1148/ryai.230514","DOIUrl":"10.1148/ryai.230514","url":null,"abstract":"<p><p>Artificial intelligence (AI) models often face performance drops after deployment to external datasets. This study evaluated the potential of a novel data augmentation framework based on generative adversarial networks (GANs) that creates synthetic patient image data for model training to improve model generalizability. Model development and external testing were performed for a given classification task, namely the detection of new fluid-attenuated inversion recovery lesions at MRI during longitudinal follow-up of patients with multiple sclerosis (MS). An internal dataset of 669 patients with MS (<i>n</i> = 3083 examinations) was used to develop an attention-based network, trained both with and without the inclusion of the GAN-based synthetic data augmentation framework. External testing was performed on 134 patients with MS from a different institution, with MR images acquired using different scanners and protocols than images used during training. Models trained using synthetic data augmentation showed a significant performance improvement when applied on external data (area under the receiver operating characteristic curve [AUC], 83.6% without synthetic data vs 93.3% with synthetic data augmentation; <i>P</i> = .03), achieving comparable results to the internal test set (AUC, 95.0%; <i>P</i> = .53), whereas models without synthetic data augmentation demonstrated a performance drop upon external testing (AUC, 93.8% on internal dataset vs 83.6% on external data; <i>P</i> = .03). Data augmentation with synthetic patient data substantially improved performance of AI models on unseen MRI data and may be extended to other clinical conditions or tasks to mitigate domain shift, limit class imbalance, and enhance the robustness of AI applications in medical imaging. <b>Keywords:</b> Brain, Brain Stem, Multiple Sclerosis, Synthetic Data Augmentation, Generative Adversarial Network <i>Supplemental material is available for this article.</i> © RSNA, 2024.</p>","PeriodicalId":29787,"journal":{"name":"Radiology-Artificial Intelligence","volume":" ","pages":"e230514"},"PeriodicalIF":8.1,"publicationDate":"2024-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11605143/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142476382","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":"Boosting Deep Learning for Interpretable Brain MRI Lesion Detection through the Integration of Radiology Report Information.","authors":"Lisong Dai, Jiayu Lei, Fenglong Ma, Zheng Sun, Haiyan Du, Houwang Zhang, Jingxuan Jiang, Jianyong Wei, Dan Wang, Guang Tan, Xinyu Song, Jinyu Zhu, Qianqian Zhao, Songtao Ai, Ai Shang, Zhaohui Li, Ya Zhang, Yuehua Li","doi":"10.1148/ryai.230520","DOIUrl":"10.1148/ryai.230520","url":null,"abstract":"<p><p>Purpose To guide the attention of a deep learning (DL) model toward MRI characteristics of brain lesions by incorporating radiology report-derived textual features to achieve interpretable lesion detection. Materials and Methods In this retrospective study, 35 282 brain MRI scans (January 2018 to June 2023) and corresponding radiology reports from center 1 were used for training, validation, and internal testing. A total of 2655 brain MRI scans (January 2022 to December 2022) from centers 2-5 were reserved for external testing. Textual features were extracted from radiology reports to guide a DL model (ReportGuidedNet) focusing on lesion characteristics. Another DL model (PlainNet) without textual features was developed for comparative analysis. Both models identified 15 conditions, including 14 diseases and normal brains. Performance of each model was assessed by calculating macro-averaged area under the receiver operating characteristic curve (ma-AUC) and micro-averaged AUC (mi-AUC). Attention maps, which visualized model attention, were assessed with a five-point Likert scale. Results ReportGuidedNet outperformed PlainNet for all diagnoses on both internal (ma-AUC, 0.93 [95% CI: 0.91, 0.95] vs 0.85 [95% CI: 0.81, 0.88]; mi-AUC, 0.93 [95% CI: 0.90, 0.95] vs 0.89 [95% CI: 0.83, 0.92]) and external (ma-AUC, 0.91 [95% CI: 0.88, 0.93] vs 0.75 [95% CI: 0.72, 0.79]; mi-AUC, 0.90 [95% CI: 0.87, 0.92] vs 0.76 [95% CI: 0.72, 0.80]) testing sets. The performance difference between internal and external testing sets was smaller for ReportGuidedNet than for PlainNet (Δma-AUC, 0.03 vs 0.10; Δmi-AUC, 0.02 vs 0.13). The Likert scale score of ReportGuidedNet was higher than that of PlainNet (mean ± SD: 2.50 ± 1.09 vs 1.32 ± 1.20; <i>P</i> < .001). Conclusion The integration of radiology report textual features improved the ability of the DL model to detect brain lesions, thereby enhancing interpretability and generalizability. <b>Keywords:</b> Deep Learning, Computer-aided Diagnosis, Knowledge-driven Model, Radiology Report, Brain MRI <i>Supplemental material is available for this article.</i> Published under a CC BY 4.0 license.</p>","PeriodicalId":29787,"journal":{"name":"Radiology-Artificial Intelligence","volume":" ","pages":"e230520"},"PeriodicalIF":8.1,"publicationDate":"2024-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11605145/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142393849","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":"Precise Image-level Localization of Intracranial Hemorrhage on Head CT Scans with Deep Learning Models Trained on Study-level Labels.","authors":"Yunan Wu, Michael Iorga, Suvarna Badhe, James Zhang, Donald R Cantrell, Elaine J Tanhehco, Nicholas Szrama, Andrew M Naidech, Michael Drakopoulos, Shamis T Hasan, Kunal M Patel, Tarek A Hijaz, Eric J Russell, Shamal Lalvani, Amit Adate, Todd B Parrish, Aggelos K Katsaggelos, Virginia B Hill","doi":"10.1148/ryai.230296","DOIUrl":"10.1148/ryai.230296","url":null,"abstract":"<p><p>Purpose To develop a highly generalizable weakly supervised model to automatically detect and localize image-level intracranial hemorrhage (ICH) by using study-level labels. Materials and Methods In this retrospective study, the proposed model was pretrained on the image-level Radiological Society of North America dataset and fine-tuned on a local dataset by using attention-based bidirectional long short-term memory networks. This local training dataset included 10 699 noncontrast head CT scans in 7469 patients, with ICH study-level labels extracted from radiology reports. Model performance was compared with that of two senior neuroradiologists on 100 random test scans using the McNemar test, and its generalizability was evaluated on an external independent dataset. Results The model achieved a positive predictive value (PPV) of 85.7% (95% CI: 84.0, 87.4) and an area under the receiver operating characteristic curve of 0.96 (95% CI: 0.96, 0.97) on the held-out local test set (<i>n</i> = 7243, 3721 female) and 89.3% (95% CI: 87.8, 90.7) and 0.96 (95% CI: 0.96, 0.97), respectively, on the external test set (<i>n</i> = 491, 178 female). For 100 randomly selected samples, the model achieved performance on par with two neuroradiologists, but with a significantly faster (<i>P</i> < .05) diagnostic time of 5.04 seconds per scan (vs 86 seconds and 22.2 seconds for the two neuroradiologists, respectively). The model's attention weights and heatmaps visually aligned with neuroradiologists' interpretations. Conclusion The proposed model demonstrated high generalizability and high PPVs, offering a valuable tool for expedited ICH detection and prioritization while reducing false-positive interruptions in radiologists' workflows. <b>Keywords:</b> Computer-Aided Diagnosis (CAD), Brain/Brain Stem, Hemorrhage, Convolutional Neural Network (CNN), Transfer Learning <i>Supplemental material is available for this article.</i> © RSNA, 2024 See also the commentary by Akinci D'Antonoli and Rudie in this issue.</p>","PeriodicalId":29787,"journal":{"name":"Radiology-Artificial Intelligence","volume":" ","pages":"e230296"},"PeriodicalIF":8.1,"publicationDate":"2024-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11605431/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142081915","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":"WAW-TACE: A Hepatocellular Carcinoma Multiphase CT Dataset with Segmentations, Radiomics Features, and Clinical Data.","authors":"Krzysztof Bartnik, Tomasz Bartczak, Mateusz Krzyziński, Krzysztof Korzeniowski, Krzysztof Lamparski, Piotr Węgrzyn, Eric Lam, Mateusz Bartkowiak, Tadeusz Wróblewski, Katarzyna Mech, Magdalena Januszewicz, Przemysław Biecek","doi":"10.1148/ryai.240296","DOIUrl":"10.1148/ryai.240296","url":null,"abstract":"","PeriodicalId":29787,"journal":{"name":"Radiology-Artificial Intelligence","volume":" ","pages":"e240296"},"PeriodicalIF":8.1,"publicationDate":"2024-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11605144/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142509377","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":"The RSNA Abdominal Traumatic Injury CT (RATIC) Dataset.","authors":"Jeffrey D Rudie, Hui-Ming Lin, Robyn L Ball, Sabeena Jalal, Luciano M Prevedello, Savvas Nicolaou, Brett S Marinelli, Adam E Flanders, Kirti Magudia, George Shih, Melissa A Davis, John Mongan, Peter D Chang, Ferco H Berger, Sebastiaan Hermans, Meng Law, Tyler Richards, Jan-Peter Grunz, Andreas Steven Kunz, Shobhit Mathur, Sandro Galea-Soler, Andrew D Chung, Saif Afat, Chin-Chi Kuo, Layal Aweidah, Ana Villanueva Campos, Arjuna Somasundaram, Felipe Antonio Sanchez Tijmes, Attaporn Jantarangkoon, Leonardo Kayat Bittencourt, Michael Brassil, Ayoub El Hajjami, Hakan Dogan, Muris Becircic, Agrahara G Bharatkumar, Eduardo Moreno Júdice de Mattos Farina, Errol Colak","doi":"10.1148/ryai.240101","DOIUrl":"10.1148/ryai.240101","url":null,"abstract":"<p><p>\u0000 <i>Supplemental material is available for this article.</i>\u0000 </p>","PeriodicalId":29787,"journal":{"name":"Radiology-Artificial Intelligence","volume":" ","pages":"e240101"},"PeriodicalIF":8.1,"publicationDate":"2024-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11605137/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142509376","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":"Breaking Ground on the Application of AI to HCC: It's All about Data.","authors":"Ryan Bitar, Julius Chapiro","doi":"10.1148/ryai.240660","DOIUrl":"10.1148/ryai.240660","url":null,"abstract":"","PeriodicalId":29787,"journal":{"name":"Radiology-Artificial Intelligence","volume":"6 6","pages":"e240660"},"PeriodicalIF":8.1,"publicationDate":"2024-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11605136/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142733009","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":"Watch Your Back! How Deep Learning Is Cracking the Real World of CT for Cervical Spine Fractures.","authors":"Riccardo Levi, Letterio S Politi","doi":"10.1148/ryai.240604","DOIUrl":"10.1148/ryai.240604","url":null,"abstract":"","PeriodicalId":29787,"journal":{"name":"Radiology-Artificial Intelligence","volume":"6 6","pages":"e240604"},"PeriodicalIF":8.1,"publicationDate":"2024-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11605139/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142733028","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":"AI-integrated Screening to Replace Double Reading of Mammograms: A Population-wide Accuracy and Feasibility Study.","authors":"Mohammad T Elhakim, Sarah W Stougaard, Ole Graumann, Mads Nielsen, Oke Gerke, Lisbet B Larsen, Benjamin S B Rasmussen","doi":"10.1148/ryai.230529","DOIUrl":"10.1148/ryai.230529","url":null,"abstract":"<p><p>Mammography screening supported by deep learning-based artificial intelligence (AI) solutions can potentially reduce workload without compromising breast cancer detection accuracy, but the site of deployment in the workflow might be crucial. This retrospective study compared three simulated AI-integrated screening scenarios with standard double reading with arbitration in a sample of 249 402 mammograms from a representative screening population. A commercial AI system replaced the first reader (scenario 1: integrated AI_first), the second reader (scenario 2: integrated AI_second), or both readers for triaging of low- and high-risk cases (scenario 3: integrated AI_triage). AI threshold values were chosen based partly on previous validation and setting the screen-read volume reduction at approximately 50% across scenarios. Detection accuracy measures were calculated. Compared with standard double reading, integrated AI_first showed no evidence of a difference in accuracy metrics except for a higher arbitration rate (+0.99%, <i>P</i> < .001). Integrated AI_second had lower sensitivity (-1.58%, <i>P</i> < .001), negative predictive value (NPV) (-0.01%, <i>P</i> < .001), and recall rate (-0.06%, <i>P</i> = .04) but a higher positive predictive value (PPV) (+0.03%, <i>P</i> < .001) and arbitration rate (+1.22%, <i>P</i> < .001). Integrated AI_triage achieved higher sensitivity (+1.33%, <i>P</i> < .001), PPV (+0.36%, <i>P</i> = .03), and NPV (+0.01%, <i>P</i> < .001) but lower arbitration rate (-0.88%, <i>P</i> < .001). Replacing one or both readers with AI seems feasible; however, the site of application in the workflow can have clinically relevant effects on accuracy and workload. <b>Keywords:</b> Mammography, Breast, Neoplasms-Primary, Screening, Epidemiology, Diagnosis, Convolutional Neural Network (CNN) <i>Supplemental material is available for this article.</i> Published under a CC BY 4.0 license.</p>","PeriodicalId":29787,"journal":{"name":"Radiology-Artificial Intelligence","volume":" ","pages":"e230529"},"PeriodicalIF":8.1,"publicationDate":"2024-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11605135/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142126863","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}