Radiology最新文献

The Ovarian-Adnexal Reporting and Data System (O-RADS) is an evidence-based clinical support system for ovarian and adnexal lesion assessment in women of average risk. The system has both US and MRI components with separate but complementary lexicons and assessment categories to assign the risk of malignancy. US is an appropriate initial imaging modality, and O-RADS US can accurately help to characterize most adnexal lesions. MRI is a valuable adjunct imaging tool to US, and O-RADS MRI can help to both confirm a benign diagnosis and accurately stratify lesions that are at risk for malignancy. This article will review the O-RADS US and MRI systems, highlight their similarities and differences, and provide an overview of the interplay between the systems. When used together, the O-RADS US and MRI systems can help to accurately diagnose benign lesions, assess the risk of malignancy in lesions suspicious for malignancy, and triage patients for optimal management.

Background Field-cycling imaging (FCI) is a new technology developed at the University of Aberdeen that measures change in T1 relaxation time constant of tissues over a range of low magnetic field strengths (0.2-200 mT) by rapidly switching between different fields during the pulse sequence. This provides new sources of contrast, including some invisible to clinical MRI scanners, and may be a useful alternative imaging modality for stroke. Purpose To test whether a prototype whole-body FCI scanner can be used to identify infarct regions in patients with subacute ischemic stroke. Materials and Methods This prospective study screened consecutive adult patients admitted to a single center stroke unit from February 2018 to March 2020 and April to December 2021. Included participants with confirmed ischemic stroke underwent FCI 1-6 days after ictus. FCI scans were obtained at four to six evolution fields between 0.2 mT and 0.2 T, with five evolution times from 5 to 546 msec. T1 maps were generated. The Wilcoxon signed-rank test was used to compare infarct region and contralateral unaffected brain, and Spearman rank correlation was used to examine associations between infarct to contralateral tissue contrast ratio and field strengths. Two independent readers blinded to clinical images rated the FCI scans. Results Nine participants (mean age, 62 years ± 16 [SD]; all male) successfully completed FCI. FCI scans below 0.2 T exhibited hyperintense T1 regions corresponding to the infarct region identified at baseline imaging, visually confirmed with 86% interrater agreement (Cohen κ = 0.69). Infarct to contralateral tissue contrast ratio increased as magnetic field decreased between 0.2 mT and 0.2 T (r[24] = -0.68; P < .001). T1 dispersion slopes differed between infarct and unaffected tissues (median, 0.23 [IQR, 0.18-0.37] vs 0.35 [IQR, 0.27-0.43]; P = .03). Conclusion Whole-brain FCI can be used to identify subacute ischemic stroke by T1 relaxation mechanisms at field strengths as low as 0.2 mT. Research Registry no. 1813 Published under a CC BY 4.0 license. Supplemental material is available for this article.

Background Lutetium 177 [¹⁷⁷Lu]Lu-PSMA-617 (¹⁷⁷Lu-PSMA-617) is a prostate-specific membrane antigen (PSMA)-targeted radioligand therapy for metastatic castration-resistant prostate cancer (mCRPC). Quantitative PSMA PET/CT analysis could provide information on ¹⁷⁷Lu-PSMA-617 treatment benefits. Purpose To explore the association between quantitative baseline gallium 68 [⁶⁸Ga]Ga-PSMA-11 (⁶⁸Ga-PSMA-11) PET/CT parameters and treatment response and outcomes in the VISION trial. Materials and Methods This was an exploratory secondary analysis of the VISION trial. Eligible participants were randomized (June 2018 to October 2019) in a 2:1 ratio to ¹⁷⁷Lu-PSMA-617 therapy (7.4 GBq every 6 weeks for up to six cycles) plus standard of care (SOC) or to SOC only. Baseline ⁶⁸Ga-PSMA-11 PET parameters, including the mean and maximum standardized uptake value (SUV_mean and SUV_max), PSMA-positive tumor volume, and tumor load, were extracted from five anatomic regions and the whole body. Associations of quantitative PET parameters with radiographic progression-free survival (rPFS), overall survival (OS), objective response rate, and prostate-specific antigen response were investigated using univariable and multivariable analyses (with treatment as the only other covariate). Outcomes were assessed in subgroups based on SUV_mean quartiles. Results Quantitative PET parameters were well balanced between study arms for the 826 participants included. The median whole-body tumor SUV_mean was 7.6 (IQR, 5.8-9.9). Whole-body tumor SUV_mean was the best predictor of ¹⁷⁷Lu-PSMA-617 efficacy, with a hazard ratio (HR) range of 0.86-1.43 for all outcomes (all P < .001). A 1-unit whole-body tumor SUV_mean increase was associated with a 12% and 10% decrease in risk of an rPFS event and death, respectively. ¹⁷⁷Lu-PSMA-617 plus SOC prolonged rPFS and OS in all SUV_mean quartiles versus SOC only, with no identifiable optimum among participants receiving ¹⁷⁷Lu-PSMA-617. Higher baseline PSMA-positive tumor volume and tumor load were associated with worse rPFS (HR range, 1.44-1.53 [P < .05] and 1.02-1.03 [P < .001], respectively) and OS (HR range, 1.36-2.12 [P < .006] and 1.04 [P < .001], respectively). Conclusion Baseline ⁶⁸Ga-PSMA-11 PET/CT whole-body tumor SUV_mean was the best predictor of ¹⁷⁷Lu-PSMA-617 efficacy in participants in the VISION trial. Improvements in rPFS and OS with ¹⁷⁷Lu-PSMA-617 plus SOC were greater among participants with higher whole-body tumor SUV_mean, with evidence for benefit at all SUV_mean levels. ClinicalTrials.gov identifier: NCT03511664 Published under a CC BY 4.0 license. Supplemental material is available for this article.

Background Large language models (LLMs) for medical applications use unknown amounts of energy, which contribute to the overall carbon footprint of the health care system. Purpose To investigate the tradeoffs between accuracy and energy use when using different LLM types and sizes for medical applications. Materials and Methods This retrospective study evaluated five different billion (B)-parameter sizes of two open-source LLMs (Meta's Llama 2, a general-purpose model, and LMSYS Org's Vicuna 1.5, a specialized fine-tuned model) using chest radiograph reports from the National Library of Medicine's Indiana University Chest X-ray Collection. Reports with missing demographic information and missing or blank files were excluded. Models were run on local compute clusters with visual computing graphic processing units. A single-task prompt explained clinical terminology and instructed each model to confirm the presence or absence of each of the 13 CheXpert disease labels. Energy use (in kilowatt-hours) was measured using an open-source tool. Accuracy was assessed with 13 CheXpert reference standard labels for diagnostic findings on chest radiographs, where overall accuracy was the mean of individual accuracies of all 13 labels. Efficiency ratios (accuracy per kilowatt-hour) were calculated for each model type and size. Results A total of 3665 chest radiograph reports were evaluated. The Vicuna 1.5 7B and 13B models had higher efficiency ratios (737.28 and 331.40, respectively) and higher overall labeling accuracy (93.83% [3438.69 of 3665 reports] and 93.65% [3432.38 of 3665 reports], respectively) than that of the Llama 2 models (7B: efficiency ratio of 13.39, accuracy of 7.91% [289.76 of 3665 reports]; 13B: efficiency ratio of 40.90, accuracy of 74.08% [2715.15 of 3665 reports]; 70B: efficiency ratio of 22.30, accuracy of 92.70% [3397.38 of 3665 reports]). Vicuna 1.5 7B had the highest efficiency ratio (737.28 vs 13.39 for Llama 2 7B). The larger Llama 2 70B model used more than seven times the energy of its 7B counterpart (4.16 kWh vs 0.59 kWh) with low overall accuracy, resulting in an efficiency ratio of only 22.30. Conclusion Smaller fine-tuned LLMs were more sustainable than larger general-purpose LLMs, using less energy without compromising accuracy, highlighting the importance of LLM selection for medical applications. © RSNA, 2024 Supplemental material is available for this article.

Background The combination of intravenous thrombolysis (IVT) with mechanical thrombectomy (MT) may have clinical benefits for patients with medium vessel occlusion. Purpose To examine whether MT combined with IVT is associated with different outcomes than MT alone in patients with acute ischemic stroke (AIS) and medium vessel occlusion. Materials and Methods This retrospective study included consecutive adult patients with AIS and medium vessel occlusion treated with MT or MT with IVT at 37 academic centers in North America, Asia, and Europe. Data were collected from September 2017 to July 2021. Propensity score matching was performed to reduce confounding. Univariable and multivariable logistic regression analyses were performed to test the association between the addition of IVT treatment and different functional and safety outcomes. Results After propensity score matching, 670 patients (median age, 75 years [IQR, 64-82 years]; 356 female) were included in the analysis; 335 underwent MT alone and 335 underwent MT with IVT. Median onset to puncture (350 vs 210 minutes, P < .001) and onset to recanalization (397 vs 273 minutes, P < .001) times were higher in the MT group than the MT with IVT group, respectively. In the univariable regression analysis, the addition of IVT was associated with higher odds of a modified Rankin Scale (mRS) score 0-2 (odds ratio [OR], 1.44; 95% CI: 1.06, 1.96; P = .019); however, this association was not observed in the multivariable analysis (OR, 1.37; 95% CI: 0.99, 1.89; P = .054). In the multivariable analysis, the addition of IVT also showed no evidence of an association with the odds of first-pass effect (OR, 1.27; 95% CI: 0.9, 1.79; P = .17), Thrombolysis in Cerebral Infarction grades 2b-3 (OR, 1.64; 95% CI: 0.99, 2.73; P = .055), mRS scores 0-1 (OR, 1.27; 95% CI: 0.91, 1.76; P = .16), mortality (OR, 0.78; 95% CI: 0.49, 1.24; P = .29), or intracranial hemorrhage (OR, 1.25; 95% CI: 0.88, 1.76; P = .21). Conclusion Adjunctive IVT may not provide benefit to MT in patients with AIS caused by distal and medium vessel occlusion. © RSNA, 2024 Supplemental material is available for this article. See also the editorial by Wojak in this issue.