Radiology-Artificial Intelligence最新文献

Purpose To evaluate the performance of the top models from the RSNA 2022 Cervical Spine Fracture Detection challenge on a clinical test dataset of both noncontrast and contrast-enhanced CT scans acquired at a level I trauma center. Materials and Methods Seven top-performing models in the RSNA 2022 Cervical Spine Fracture Detection challenge were retrospectively evaluated on a clinical test set of 1828 CT scans (from 1829 series: 130 positive for fracture, 1699 negative for fracture; 1308 noncontrast, 521 contrast enhanced) from 1779 patients (mean age, 55.8 years ± 22.1 [SD]; 1154 [64.9%] male patients). Scans were acquired without exclusion criteria over 1 year (January-December 2022) from the emergency department of a neurosurgical and level I trauma center. Model performance was assessed using area under the receiver operating characteristic curve (AUC), sensitivity, and specificity. False-positive and false-negative cases were further analyzed by a neuroradiologist. Results Although all seven models showed decreased performance on the clinical test set compared with the challenge dataset, the models maintained high performances. On noncontrast CT scans, the models achieved a mean AUC of 0.89 (range: 0.79-0.92), sensitivity of 67.0% (range: 30.9%-80.0%), and specificity of 92.9% (range: 82.1%-99.0%). On contrast-enhanced CT scans, the models had a mean AUC of 0.88 (range: 0.76-0.94), sensitivity of 81.9% (range: 42.7%-100.0%), and specificity of 72.1% (range: 16.4%-92.8%). The models identified 10 fractures missed by radiologists. False-positive cases were more common in contrast-enhanced scans and observed in patients with degenerative changes on noncontrast scans, while false-negative cases were often associated with degenerative changes and osteopenia. Conclusion The winning models from the 2022 RSNA AI Challenge demonstrated a high performance for cervical spine fracture detection on a clinical test dataset, warranting further evaluation for their use as clinical support tools. Keywords: Feature Detection, Supervised Learning, Convolutional Neural Network (CNN), Genetic Algorithms, CT, Spine, Technology Assessment, Head/Neck Supplemental material is available for this article. © RSNA, 2024 See also commentary by Levi and Politi in this issue.

Purpose To test the performance of a transformer-based model when manipulating pretraining weights, dataset size, and input size and comparing the best model with the reference standard and state-of-the-art models for a resting-state functional (rs-fMRI) fetal brain extraction task. Materials and Methods An internal retrospective dataset (172 fetuses, 519 images; collected 2018-2022) was used to investigate influence of dataset size, pretraining approaches, and image input size on Swin-U-Net transformer (UNETR) and UNETR models. The internal and external (131 fetuses, 561 images) datasets were used to cross-validate and to assess generalization capability of the best model versus state-of-the-art models on different scanner types and number of gestational weeks (GWs). The Dice similarity coefficient (DSC) and the balanced average Hausdorff distance (BAHD) were used as segmentation performance metrics. Generalized equation estimation multifactorial models were used to assess significant model and interaction effects of interest. Results The Swin-UNETR model was not affected by the pretraining approach and dataset size and performed best with the mean dataset image size, with a mean DSC of 0.92 and BAHD of 0.097. Swin-UNETR was not affected by scanner type. Generalization results on the internal dataset showed that Swin-UNETR had lower performance compared with the reference standard models and comparable performance on the external dataset. Cross-validation on internal and external test sets demonstrated better and comparable performance of Swin-UNETR versus convolutional neural network architectures during the late-fetal period (GWs > 25) but lower performance during the midfetal period (GWs ≤ 25). Conclusion Swin-UNTER showed flexibility in dealing with smaller datasets, regardless of pretraining approaches. For fetal brain extraction from rs-fMR images, Swin-UNTER showed comparable performance with that of reference standard models during the late-fetal period and lower performance during the early GW period. Keywords: Transformers, CNN, Medical Imaging Segmentation, MRI, Dataset Size, Input Size, Transfer Learning Supplemental material is available for this article. © RSNA, 2024.

Purpose To assess the prognostic value of a deep learning-based chest radiographic age (hereafter, CXR-Age) model in a large external test cohort of Asian individuals. Materials and Methods This single-center, retrospective study included chest radiographs from consecutive, asymptomatic Asian individuals aged 50-80 years who underwent health checkups between January 2004 and June 2018. This study performed a dedicated external test of a previously developed CXR-Age model, which predicts an age adjusted based on the risk of all-cause mortality. Adjusted hazard ratios (HRs) of CXR-Age for all-cause, cardiovascular, lung cancer, and respiratory disease mortality were assessed using multivariable Cox or Fine-Gray models, and their added values were evaluated by likelihood ratio tests. Results A total of 36 924 individuals (mean chronological age, 58 years ± 7 [SD]; CXR-Age, 60 years ± 5; 22 352 male) were included. During a median follow-up of 11.0 years, 1250 individuals (3.4%) died, including 153 cardiovascular (0.4%), 166 lung cancer (0.4%), and 98 respiratory (0.3%) deaths. CXR-Age was a significant risk factor for all-cause (adjusted HR at chronological age of 50 years, 1.03; at 60 years, 1.05; at 70 years, 1.07), cardiovascular (adjusted HR, 1.11), lung cancer (adjusted HR for individuals who formerly smoked, 1.12; for those who currently smoke, 1.05), and respiratory disease (adjusted HR, 1.12) mortality (P < .05 for all). The likelihood ratio test demonstrated added prognostic value of CXR-Age to clinical factors, including chronological age for all outcomes (P < .001 for all). Conclusion Deep learning-based chest radiographic age was associated with various survival outcomes and had added value to clinical factors in asymptomatic Asian individuals, suggesting its generalizability. Keywords: Conventional Radiography, Thorax, Heart, Lung, Mediastinum, Outcomes Analysis, Quantification, Prognosis, Convolutional Neural Network (CNN) Supplemental material is available for this article. © RSNA, 2024 See also the commentary by Adams and Bressem in this issue.

Purpose To determine whether the unsupervised domain adaptation (UDA) method with generated images improves the performance of a supervised learning (SL) model for prostate cancer (PCa) detection using multisite biparametric (bp) MRI datasets. Materials and Methods This retrospective study included data from 5150 patients (14 191 samples) collected across nine different imaging centers. A novel UDA method using a unified generative model was developed for PCa detection using multisite bpMRI datasets. This method translates diffusion-weighted imaging (DWI) acquisitions, including apparent diffusion coefficient (ADC) and individual diffusion-weighted (DW) images acquired using various b values, to align with the style of images acquired using b values recommended by Prostate Imaging Reporting and Data System (PI-RADS) guidelines. The generated ADC and DW images replace the original images for PCa detection. An independent set of 1692 test cases (2393 samples) was used for evaluation. The area under the receiver operating characteristic curve (AUC) was used as the primary metric, and statistical analysis was performed via bootstrapping. Results For all test cases, the AUC values for baseline SL and UDA methods were 0.73 and 0.79 (P < .001), respectively, for PCa lesions with PI-RADS score of 3 or greater and 0.77 and 0.80 (P < .001) for lesions with PI-RADS scores of 4 or greater. In the 361 test cases under the most unfavorable image acquisition setting, the AUC values for baseline SL and UDA were 0.49 and 0.76 (P < .001) for lesions with PI-RADS scores of 3 or greater and 0.50 and 0.77 (P < .001) for lesions with PI-RADS scores of 4 or greater. Conclusion UDA with generated images improved the performance of SL methods in PCa lesion detection across multisite datasets with various b values, especially for images acquired with significant deviations from the PI-RADS-recommended DWI protocol (eg, with an extremely high b value). Keywords: Prostate Cancer Detection, Multisite, Unsupervised Domain Adaptation, Diffusion-weighted Imaging, b Value Supplemental material is available for this article. © RSNA, 2024.

Purpose To evaluate nnU-Net-based segmentation models for automated delineation of medulloblastoma tumors on multi-institutional MRI scans. Materials and Methods This retrospective study included 78 pediatric patients (52 male, 26 female), with ages ranging from 2 to 18 years, with medulloblastomas, from three different sites (28 from hospital A, 18 from hospital B, and 32 from hospital C), who had data available from three clinical MRI protocols (gadolinium-enhanced T1-weighted, T2-weighted, and fluid-attenuated inversion recovery). The scans were retrospectively collected from the year 2000 until May 2019. Reference standard annotations of the tumor habitat, including enhancing tumor, edema, and cystic core plus nonenhancing tumor subcompartments, were performed by two experienced neuroradiologists. Preprocessing included registration to age-appropriate atlases, skull stripping, bias correction, and intensity matching. The two models were trained as follows: (a) the transfer learning nnU-Net model was pretrained on an adult glioma cohort (n = 484) and fine-tuned on medulloblastoma studies using Models Genesis and (b) the direct deep learning nnU-Net model was trained directly on the medulloblastoma datasets, across fivefold cross-validation. Model robustness was evaluated on the three datasets when using different combinations of training and test sets, with data from two sites at a time used for training and data from the third site used for testing. Results Analysis on the three test sites yielded Dice scores of 0.81, 0.86, and 0.86 and 0.80, 0.86, and 0.85 for tumor habitat; 0.68, 0.84, and 0.77 and 0.67, 0.83, and 0.76 for enhancing tumor; 0.56, 0.71, and 0.69 and 0.56, 0.71, and 0.70 for edema; and 0.32, 0.48, and 0.43 and 0.29, 0.44, and 0.41 for cystic core plus nonenhancing tumor for the transfer learning and direct nnU-Net models, respectively. The models were largely robust to site-specific variations. Conclusion nnU-Net segmentation models hold promise for accurate, robust automated delineation of medulloblastoma tumor subcompartments, potentially leading to more effective radiation therapy planning in pediatric medulloblastoma. Keywords: Pediatrics, MR Imaging, Segmentation, Transfer Learning, Medulloblastoma, nnU-Net, MRI Supplemental material is available for this article. © RSNA, 2024 See also the commentary by Rudie and Correia de Verdier in this issue.

The diagnostic performance of an artificial intelligence (AI) clinical decision support solution for acute intracranial hemorrhage (ICH) detection was assessed in a large teleradiology practice. The impact on radiologist read times and system efficiency was also quantified. A total of 61 704 consecutive noncontrast head CT examinations were retrospectively evaluated. System performance was calculated along with mean and median read times for CT studies obtained before (baseline, pre-AI period; August 2021 to May 2022) and after (post-AI period; January 2023 to February 2024) AI implementation. The AI solution had a sensitivity of 75.6%, specificity of 92.1%, accuracy of 91.7%, prevalence of 2.70%, and positive predictive value of 21.1%. Of the 56 745 post-AI CT scans with no bleed identified by a radiologist, examinations falsely flagged as suspected ICH by the AI solution (n = 4464) took an average of 9 minutes 40 seconds (median, 8 minutes 7 seconds) to interpret as compared with 8 minutes 25 seconds (median, 6 minutes 48 seconds) for unremarkable CT scans before AI (n = 49 007) (P < .001) and 8 minutes 38 seconds (median, 6 minutes 53 seconds) after AI when ICH was not suspected by the AI solution (n = 52 281) (P < .001). CT scans with no bleed identified by the AI but reported as positive for ICH by the radiologist (n = 384) took an average of 14 minutes 23 seconds (median, 13 minutes 35 seconds) to interpret as compared with 13 minutes 34 seconds (median, 12 minutes 30 seconds) for CT scans correctly reported as a bleed by the AI (n = 1192) (P = .04). With lengthened read times for falsely flagged examinations, system inefficiencies may outweigh the potential benefits of using the tool in a high volume, low prevalence environment. Keywords: Artificial Intelligence, Intracranial Hemorrhage, Read Time, Report Turnaround Time, System Efficiency Supplemental material is available for this article. © RSNA, 2024.

Purpose To develop and evaluate a publicly available deep learning model for segmenting and classifying cardiac implantable electronic devices (CIEDs) on Digital Imaging and Communications in Medicine (DICOM) and smartphone-based chest radiographs. Materials and Methods This institutional review board-approved retrospective study included patients with implantable pacemakers, cardioverter defibrillators, cardiac resynchronization therapy devices, and cardiac monitors who underwent chest radiography between January 2012 and January 2022. A U-Net model with a ResNet-50 backbone was created to classify CIEDs on DICOM and smartphone images. Using 2321 chest radiographs in 897 patients (median age, 76 years [range, 18-96 years]; 625 male, 272 female), CIEDs were categorized into four manufacturers, 27 models, and one "other" category. Five smartphones were used to acquire 11 072 images. Performance was reported using the Dice coefficient on the validation set for segmentation or balanced accuracy on the test set for manufacturer and model classification, respectively. Results The segmentation tool achieved a mean Dice coefficient of 0.936 (IQR: 0.890-0.958). The model had an accuracy of 94.36% (95% CI: 90.93%, 96.84%; 251 of 266) for CIED manufacturer classification and 84.21% (95% CI: 79.31%, 88.30%; 224 of 266) for CIED model classification. Conclusion The proposed deep learning model, trained on both traditional DICOM and smartphone images, showed high accuracy for segmentation and classification of CIEDs on chest radiographs. Keywords: Conventional Radiography, Segmentation Supplemental material is available for this article. © RSNA, 2024 See also the commentary by Júdice de Mattos Farina and Celi in this issue.