Signal Intensity Trajectories Clustering for Liver Vasculature Segmentation and Labeling (LiVaS) on Contrast-Enhanced MR Images: A Feasibility Pilot Study.

This study aims to develop a semiautomated pipeline and user interface (LiVaS) for rapid segmentation and labeling of MRI liver vasculature and evaluate its time efficiency and accuracy against manual reference standard. Retrospective feasibility pilot study. Liver MR images from different scanners from 36 patients were included, and 4/36 patients were randomly selected for manual segmentation as referenced standard. The liver was segmented in each contrast phase and masks registered to the pre-contrast segmentation. Voxel-wise signal trajectories were clustered using the k-means algorithm. Voxel clusters that best segment the liver vessels were selected and labeled by three independent radiologists and a research scientist using LiVaS. Segmentation times were compared using a paired-sample t-test on log-transformed data. The agreement was analyzed qualitatively and quantitatively using DSC for hepatic and portal vein segmentations. The mean segmentation time among four readers was significantly shorter than manual (3.6 ± 1.4 vs. 70.0 ± 29.2 min; p < 0.001), even when using a higher number of clusters to enhance accuracy. The DSC for portal and hepatic veins reached up to 0.69 and 0.70, respectively. LiVaS segmentations were overall of good quality, with variations in performance related to the presence/severity of liver disease, acquisition timing, and image quality. Our semi-automated pipeline was robust to different MRI vendors in producing segmentation and labeling of liver vasculature in agreement with expert manual annotations, with significantly higher time efficiency. LiVaS could facilitate the creation of large, annotated datasets for training and validation of neural networks for automated MRI liver vascularity segmentation. HIGHLIGHTS: Key Finding: In this pilot feasibility study, our semiautomated pipeline for segmentation of liver vascularity (LiVaS) on MR images produced segmentations with simultaneous labeling of portal and hepatic veins in good agreement with the manual reference standard but at significantly shorter times (mean LiVaS 3.6 ± 1.4 vs. mean manual 70.0 ± 29.2 min; p < 0.001). Importance: LiVaS was robust in producing liver MRI vascular segmentations across images from different scanners in agreement with expert manual annotations, with significant ly higher time efficiency, and therefore potential scalability.

Ultrasound Attenuation Imaging vs MRI-PDFF, Echogenicity and Liver Function for Assessing Degree of Steatosis in NAFLD and Non-NAFLD Patients.

ABSTRACT: Nonalcoholic fatty liver disease (NAFLD) is a primary cause of parenchymal liver disease globally. There are currently several methods available to test the degree of steatosis in NAFLD patients, but all have drawbacks that limit their use.The objective of this study is to determine if a new technique, ultrasound (US) attenuation imaging (ATI), correlates with magnetic resonance proton density fat fraction imaging and hepatic echogenicity as seen on gray scale US imaging.Fifty-four patients were recruited at the University of Washington Medical Center from individuals who had already been scheduled for hepatic US or magnetic resonance imaging (MRI). All participants then underwent both hepatic MRI proton density fat fraction and US. Ultrasound images were then evaluated using ATI with 2 observers who individually determined relative grayscale echogenicity.Analysis showed positive correlation between ATI- and MRI-determined fat percentage in the case group (Spearman correlation: 0.50; P = 0.015). Furthermore, participants with NAFLD tended to have a higher ATI than controls (median: 0.70 vs 0.54 dB/cm/MHz; P < 0.001).This study demonstrates that US ATI combined with grayscale imaging is an effective way of assessing the degree of steatosis in patients with moderate to severe NAFLD.

Multi-Omic Biomarkers Improve Indeterminate Pulmonary Nodule Malignancy Risk Assessment.

The clinical management of patients with indeterminate pulmonary nodules is associated with unintended harm to patients and better methods are required to more precisely quantify lung cancer risk in this group. Here, we combine multiple noninvasive approaches to more accurately identify lung cancer in indeterminate pulmonary nodules. We analyzed 94 quantitative radiomic imaging features and 41 qualitative semantic imaging variables with molecular biomarkers from blood derived from an antibody-based microarray platform that determines protein, cancer-specific glycan, and autoantibody-antigen complex content with high sensitivity. From these datasets, we created a PSR (plasma, semantic, radiomic) risk prediction model comprising nine blood-based and imaging biomarkers with an area under the receiver operating curve (AUROC) of 0.964 that when tested in a second, independent cohort yielded an AUROC of 0.846. Incorporating known clinical risk factors (age, gender, and smoking pack years) for lung cancer into the PSR model improved the AUROC to 0.897 in the second cohort and was more accurate than a well-characterized clinical risk prediction model (AUROC = 0.802). Our findings support the use of a multi-omics approach to guide the clinical management of indeterminate pulmonary nodules.

CT perfusion as a potential biomarker for pancreatic ductal adenocarcinoma during routine staging and restaging.

PURPOSE: To evaluate the significance of CT perfusion parameters predicting response to neoadjuvant therapy in patients with pancreatic ductal adenocarcinoma (PDAC). MATERIALS AND METHODS: Seventy patients with PDAC prospectively had CT perfusion acquisition incorporated into baseline multiphase staging CT. Twenty-eight who were naïve to therapy were retained for further investigation. Perfusion was performed 5-42.5 s after contrast, followed by parenchymal and portal venous phases. Blood flow (BF), blood volume (BV), and permeability surface area product (PS) were calculated using deconvolution algorithms. Patients were categorized as responders or non-responders per RECIST 1.1. Perfusion variables with AUC ≥ 0.70 in differentiating responders from non-responders were retained. Logistic regression was used to assess associations between baseline perfusion variables and response. RESULTS: 18 of 28 patients showed favorable response to therapy. Baseline heterogeneity variables in tumor max ROI were higher in non-responders than responders [median BF coefficient of variation (CV) 0.91 vs. 0.51 respectively, odds ratio (OR) 6.8 per one standard deviation (1-SD) increase, P = 0.047; median PS CV 1.6 vs. 0.68, OR 3.9 per 1-SD increase, P = 0.047; and median BV CV 0.75 vs. 0.54, OR = 4.0 per 1-SD increase, P = 0.047]. Baseline BV mean in tumor center was lower in non-responders than responders (median BV mean: 0.74 vs. 2.9 ml/100 g respectively, OR 0.28 per 1-SD increase, P = 0.047). CONCLUSION: For patients with PDAC receiving neoadjuvant therapy, lower and more heterogeneous perfusion parameters correlated with an unfavorable response to therapy. Such quantitative information can be acquired utilizing a comprehensive protocol interleaving perfusion CT acquisition with standard of care multiphase CT scans using a single contrast injection, which could be used to identify surgical candidates and predict outcome.

Accelerated Breast Diffusion-weighted Imaging Using Multiband Sensitivity Encoding with the CAIPIRINHA Method: Clinical Experience at 3 T.

Purpose To examine the clinical value of multiband (MB) sensitivity encoding (SENSE)-accelerated diffusion-weighted imaging (DWI) for breast imaging by performing quantitative and qualitative comparisons with conventional diffusion-weighted echo-planar imaging, or conventional DWI (cDWI). Materials and Methods In this prospective study (ClinicalTrials.gov identifier NCT03607552), women with breast cancer were recruited from July 2018 to July 2019 to undergo additional MB SENSE DWI during clinical 3-T breast MRI examinations. The cDWI and MB SENSE DWI acquisitions were assessed both quantitatively and qualitatively. Regions of interest were defined for tumorous and normal tissue, and the tumor apparent diffusion coefficient (ADC), contrast-to-noise ratio (CNR), and signal index (SI) were calculated for both DWI methods. Three readers independently reviewed the two acquisitions side by side and provided relative image quality scores. Tumor ADC, CNR, and SI measures were compared between cDWI and MB SENSE DWI acquisitions by using a paired t test, and reader preferences were evaluated by using the sign test. Results The study included 38 women (median age, 48 years; range, 28-83 years). Overall agreement was good between cDWI and MB SENSE DWI tumor ADC measures (intraclass correlation coefficient, 0.87 [95% CI: 0.75, 0.94]), and no differences were evident in the ADC (median, 0.93 × 10-3 mm2/sec vs 0.87 ×10-3 mm2/sec; P = .50), CNR (2.2 vs 2.3; P = .17), or SI (9.2 vs 9.2; P = .23) measurements. The image quality of cDWI and MB SENSE DWI acquisitions were considered equal for 51% of images (58 of 114), whereas MB SENSE DWI was preferred more often than cDWI (37% [42 of 114] vs 12% [14 of 114]; P < .001). The preference for MB SENSE DWI was most often attributed to better fat suppression. Conclusion MB SENSE can be used to accelerate breast DWI acquisition times without compromising the image quality or the fidelity of quantitative ADC measurements. Keywords: MR-Diffusion-weighted Imaging, Breast, Comparative Studies, Technology Assessment Clinical trial registration no. NCT03607552 © RSNA, 2022.

Magnetic Resonance Prediction of Lung Maturity in Fetuses With Congenital Diaphragmatic Hernia.

ABSTRACT: To determine if lung to liver MR T2 signal ratio is predictive of neonatal outcome in fetuses with congenital diaphragmatic hernia (CDH).After Interal Review Board approval, the PACS systems at the University of Washington and University of Utah were searched for cases having an in utero fetal MR examination diagnostic of CDH. Inclusion criteria were at least 1 prior ultrasound demonstrating a CDH and an MR obtained within 1 week of that prior ultrasound.A total of 69 patients from the University of Utah and 13 from the University of Washington satisfied the inclusion criteria for a total of 82. After adjusting for gestational age and contralateral lung volume, there was little apparent association between contralateral lung to liver MR T2 signal and 5-minute Apgar score and neonatal mortality When considering neonatal Apgar and mortality, increasing contralateral lung volume was significantly associated with lower risk (hazard ratio, 0.40 per doubling; 95% confidence interval, 0.24-0.69; P = 0.001) as expected.Our data demonstrate that the lung to liver MR signal ratio was not predictive of outcome. The measurement of contralateral lung area, and gestational age at the time of the examination (time of diagnosis) are still the best predictors of poor neonatal outcome.

Virtual Unenhanced Dual-Energy CT Images Obtained with a Multimaterial Decomposition Algorithm: Diagnostic Value for Renal Mass and Urinary Stone Evaluation.

Background Virtual unenhanced (VUE) images obtained by using a dual-energy CT (DECT) multimaterial decomposition algorithm hold promise for diagnostic use in the abdomen in lieu of true unenhanced (TUE) images. Purpose To assess VUE images obtained from a DECT multimaterial decomposition algorithm in patients undergoing renal mass and urinary stone evaluation. Materials and Methods In this retrospective Health Insurance Portability and Accountability Act-compliant study, DECT was performed in patients undergoing evaluation for renal mass or urinary stone. VUE images were compared quantitatively to TUE images and qualitatively assessed by four independent radiologists. Differences in attenuation between VUE and TUE images were summarized by using 95% limits of agreement. Diagnostic performance in urinary stone detection was summarized by using area under the receiver operating characteristic curve, sensitivity, and specificity. Results A total of 221 patients (mean age ± standard deviation, 61 years ± 14; 129 men) with 273 renal masses were evaluated. Differences in renal mass attenuation between VUE and TUE images were within 3 HU for both enhancing masses (95% limits of agreement, -3.1 HU to 2.7 HU) and nonenhancing cysts (95% limits of agreement, -2.9 HU to 2.5 HU). Renal mass classification as enhancing mass versus nonenhancing cyst did not change (reclassification rate of enhancing masses, 0% [0 of 78]; 95% CI: 0, 5; reclassification rate of nonenhancing cysts, 0% [0 of 193]; 95% CI: 0, 2) with use of VUE in lieu of TUE images. Among 166 urinary stones evaluated, diagnostic performance of VUE images for stone detection was lower compared with that of TUE images (area under the receiver operating characteristic curve, 0.79 [95% CI: 0.73, 0.84] vs 0.93 [95% CI: 0.91, 0.95]; P < .001) due to reduced sensitivity of VUE for detection of stones 3 mm in diameter or less compared with those greater than 3 mm (sensitivity, 23% [25 of 108; 95% CI: 12, 40] vs 88% [126 of 144; 95% CI: 77, 94]; P < .001). Conclusion Compared with true unenhanced images, virtual unenhanced (VUE) images were unlikely to change renal mass classification as enhancing mass versus nonenhancing cyst. Diagnostic performance of VUE images remained suboptimal for urinary stone detection due to subtraction of stones 3 mm or less in diameter. © RSNA, 2021 Online supplemental material is available for this article. See also the editorial by Sosna in this issue.

Breast Cancer Conspicuity on Computed Versus Acquired High b-Value Diffusion-Weighted MRI.

RATIONALE AND OBJECTIVES: On unenhanced diffusion-weighted imaging (DWI), computing or synthesizing high b-value images from lower b-value acquisitions can enhance breast cancer visibility. This study aimed to evaluate relative lesion conspicuity on computed versus acquired diffusion-weighted images and investigate clinical characteristics influencing optimal b-values. MATERIALS AND METHODS: Women with newly diagnosed breast cancer were prospectively enrolled and underwent 3T breast MRI with DWI. Lesion contrast-to-noise ratio (CNR) was measured across a range of b-values (0-2500 s/mm2) for computed and acquired DWI. Three readers independently compared lesion visibility between computed and acquired DWI and selected the optimal b-value. Computed versus acquired DWI was compared quantitatively based on CNR by paired t-test and qualitatively based on reader preference using a sign test. Optimal b-values by qualitative and quantitative assessment were compared by paired t-test, and associations with clinical characteristics were assessed by Wilcoxon rank sum test. RESULTS: The study included 30 women (median age, 48 years); 28 with invasive carcinoma, 2 DCIS. Lesion CNR was higher on acquired versus computed images (p = 0.018), while lesion visibility by reader assessment was not different (p = 0.36). Optimal b-values selected by readers (mean, b = 1411 ± 383 s/mm2) were slightly higher than those based on peak CNR (b = 1233 ± 463 s/mm2, p = 0.023), and were higher for younger (≤50 years) versus older women (p = 0.002) and dense versus nondense breasts (p = 0.015). CONCLUSION: Lesion CNR on computed high b-value images was slightly reduced versus acquired images, but our study suggests that this did not significantly impact lesion visibility. Computing high b-value images offers extra flexibility to adjust b-value during interpretation.

ACR Appropriateness Criteria Facilitate Judicious Use of CT Angiography for Stroke Workup in the Emergency Department.

PURPOSE: CT angiography (CTA) of the head and neck plays a key role in the evaluation of patients with stroke in the emergency department, particularly to evaluate for large-vessel occlusion and consider patients for mechanical thrombectomy. However, CTA is not always indicated and has various costs. The aim of this study was to examine the correlation between ACR imaging appropriateness recommendations and findings explanatory for stroke on CTA of the head and neck. METHODS: CTA head and neck examinations of emergency department patients performed to evaluate for stroke between January 1, 2019, and June 30, 2019, were retrospectively reviewed. Patients with previously known acute infarct, with primary concern for or discovered hemorrhage, or being evaluated for trauma were excluded. Presenting symptoms were classified on the basis of the most recent ACR Appropriateness Criteria, into three categories: "usually appropriate" (UA), "may be appropriate" (MBA), or "usually not appropriate" (UNA). The percentage of examinations with findings explanatory for stroke were compared among the three groups using the Fisher exact test. RESULTS: A total of 707 cases were reviewed, of which 317 were included; 199 fell into the UA category, 38 MBA, and 80 UNA. For UA, 57 of 199 CTAs (29%) had findings explanatory for stroke, compared with 1 of 38 (3%) in the MBA group (difference, 26%; 95% confidence interval, 13%-34%; P < .001) and 1 of 80 (1%) in the UNA group (difference, 27%; 95% confidence interval, 19%-35%; P < .001). CONCLUSIONS: ACR Appropriateness Criteria recommendations correlate with imaging findings explanatory for stroke and may guide the judicious use of CTA for suspected stroke workup.

Emergency Radiology: An Underappreciated Source of Liability Risk.

PURPOSE: While several studies analyze radiology malpractice lawsuits, none specifically examines the site of service. The purpose of this study is to estimate the relative likelihood of a lawsuit arising from a radiology study performed in emergency (ED), inpatient (IP) and outpatient (OP) settings. METHODS: Referrals from a malpractice review consulting company over a six year period were compared to the 2016 Medicare Part B file and stratified by site of service. The proportion of exams for each site of service was estimated, and using absolute differences in proportions and odds ratios (ORs), differences in the place of service were calculated. RESULTS: The Cleareview cohort contained 25 (17%) IP, 56 (38%) OP, and 68 (46%) ED exams. In 2016, Medicare assigned benefits for 27,009,053 (20%) IP, 84,075,848 (62%) OP and 23,964,794 (18%) ED exams. The ORs (Cleareview: Medicare) of the ED to IP, OP, and IP+OP were 3.07 (95% CI: 1.56-6.03), 4.26 (95% CI: 2.76-6.59), 3.89 (95% CI: 2.60-5.83), respectively. By contrast, the OR for IP:OP between Cleareview and Medicare was not significantly different than 1 (OR: 1.39, 95% CI: 0.68-2.83, P = .38). DISCUSSION: Radiological studies performed in the ED accounted for a disproportionate number of liability claims against radiologists. Further study is warranted to confirm this finding with a more robust data set.