Search

Enlarged perivascular spaces (EPVS), specifically in stroke patients, has been shown to strongly correlate with other measures of small vessel disease and cognitive impairment at 1 year follow-up. Typical grading of EPVS is often challenging and time consuming and is usually based on a subjective visual rating scale. The purpose of the current study was to develop an interpretable, 3D neural network for grading enlarged perivascular spaces (EPVS) severity at the level of the basal ganglia using clinical-grade imaging in a heterogenous acute stroke cohort, in the context of total cerebral small vessel disease (CSVD) burden. T2-weighted images from a retrospective cohort of 262 acute stroke patients, collected in 2015 from 5 regional medical centers, were used for analyses. Patients were given a label of 0 for none-to-mild EPVS (< 10) and 1 for moderate-to-severe EPVS (≥ 10). A three-dimensional residual network of 152 layers (3D-ResNet-152) was created to predict EPVS severity and 3D gradient class activation mapping (3DGradCAM) was used for visual interpretation of results. Our model achieved an accuracy 0.897 and area-under-the-curve of 0.879 on a hold-out test set of 15% of the total cohort (n = 39). 3DGradCAM showed areas of focus that were in physiologically valid locations, including other prevalent areas for EPVS. These maps also suggested that distribution of class activation values is indicative of the confidence in the model's decision. Potential clinical implications of our results include: (1) support for feasibility of automated of EPVS scoring using clinical-grade neuroimaging data, potentially alleviating rater subjectivity and improving confidence of visual rating scales, and (2) demonstration that explainable models are critical for clinical translation., (© 2022. The Author(s).)

Rationale and Objectives: Following state and institutional guidelines, our Radiology department launched the "Recover Wisely" for all nonurgent radiology care on May 4, 2020. Our objective is to report our practice implementation and experience of COVID-19 recovery during the resumption of routine imaging at a tertiary academic medical center., Materials and Methods: We used the SQUIRE 2.0 guidelines for this practice implementation. Recover Wisely focused on a data driven, strategic rescheduling and redesigning patient flow process. We used scheduling simulations and meticulous monitoring and control of outpatient medical imaging volumes to achieve a linear restoration to our pre-COVID imaging studies. We had a tiered plan to address the backlog of rescheduled patients with gradual opening of our imaging facilities, while maintaining broad communication with our patients and referring clinicians., Results: Recover Wisely followed our anticipated linear modeling. Considering the last 10 weeks in the recovery, outpatient growth was linear with an increase of approximately 172 cases per week, (R 2 =0.97). We achieved an overall recovery of 102% in week 10, as compared to average weekly pre-COVID outpatient volumes. The modalities recovered as follows in outpatient volumes: CT (113%), MRI (101%), nuclear medicine including PET (138%), mammograms (97%), ultrasound (99%) and interventional radiology (106%). When compared to identical 2019 calendar weeks (May 4, 2020-July 10, 2020), the total 2020 radiology volume was 11% reduced from the 2019 volume. The reduction in total weighted relative value units was 8% in this time period, as compared to 2019., Conclusion: Our department utilized a data-driven, team approach based on our guiding principles to "Recover Wisely." We created and implemented a methodology that achieved a linear increase in outpatient studies over a 10-week recovery period., Competing Interests: Conflicts of Interests AV: R01 NIH/NINDS NS103824-01; R01 NIH/NINDS NS100417; NIH/NINDS 1U01NS100699; NIH/NINDS U01NS110772; Imaging Core Lab, ENDOLOW Trial, Cerenovus, Johnson & Johnson; Human centered design grant, ACR Innovation Fund. FR: Director of Medical Affairs at Imagia. MM: RSNA BOD, ACR BOC, Nonfinancial support from GE research agreement, Personal fees from Elsevier. JA, BA, JH, SC, KC, SK, MG, AM, AB, SB, EE, JS, GU: No conflicts to disclose, (Copyright © 2020 The Association of University Radiologists. Published by Elsevier Inc. All rights reserved.)

OBJECTIVE. This article provides comments from a small group of highly qualified reviewers of the American Journal of Roentgenology ( AJR ) regarding their approach to assessing manuscripts. The objective is to educate authors about the issues to which reviewers particularly attend and about errors that will decrease the likelihood of publication. CONCLUSION. By following the advice provided in this article, authors should be able to compose better manuscripts and reviewers should be able to generate better reviews.

Purpose: Texture analysis (TA) can quantify variations in surface intensity or patterns, including some that are imperceptible to the human visual system. The purpose of this study was to determine the diagnostic accuracy of radiomic based filtration-histogram TA to differentiate high-grade from low-grade gliomas by assessing tumor heterogeneity., Methods: Patients with a histopathological diagnosis of glioma and preoperative 3T MRI imaging were included in this retrospective study. A region of interest was manually delineated on post-contrast T1 images. TA was performed using commercially available research software. The histogram parameters including mean, standard deviation, entropy, mean of the positive pixels, skewness, and kurtosis were analyzed at spatial scaling factors ranging from 0 to 6 mm. The parameters were correlated with WHO glioma grade using Spearman correlation. Areas under the curve (AUC) were calculated using ROC curve analysis to distinguish tumor grades., Results: Of a total of 94 patients, 14 had low-grade gliomas and 80 had high-grade gliomas. Mean, SD, MPP, entropy and kurtosis each showed significant differences between glioma grades for different spatial scaling filters. Low and high-grade gliomas were best-discriminated using mean of 2 mm fine texture scale, with a sensitivity and specificity of 93% and 86% (AUC of 0.90)., Conclusions: Quantitative measurement of heterogeneity using TA can discriminate high versus low-grade gliomas. Radiomic data of texture features can provide complementary diagnostic information for gliomas.

Purpose: The diagnosis of intracranial tuberculomas is often challenging. Our purpose is to describe the most common metabolic patterns of tuberculomas by MR spectroscopy (MRS) with emphasis on potential specific markers., Methods: Single-voxel MRS short echo time was performed in 13 cases of tuberculomas proven by histology and/or response to anti-mycobacterial therapy. For comparison MRS was also performed in 19 biopsy-proven malignant tumors (13 high-grade gliomas and six metastasis). Presence of metabolic peaks was assessed visually and categorical variables between groups were compared using chi-square. Metabolite ratios were compared using Mann-Whitney test and diagnostic accuracy of the metabolite ratios was compared using receiver-operating characteristic (ROC) curves analysis., Results: Spectroscopic peaks representing lipids and glutamate/glutamine (Glx) as well as a peak at ∼3.8 ppm were well defined in 77% (10/13), 77% (10/13) and 69% (nine of 13) of tuberculomas, respectively. Lipid and Glx peaks were also present in most of the malignant lesions, 79% (15/19) and 74% (14/19) respectively. However, a peak at ∼3.8 ppm was present in only 10% (two of 19) of the tumor cases (p < 0.001). Higher Cho/Cr and mI/Cr ratios helped discriminate malignant lesions with an area under the ROC curve of 0.86 (SE: 0.078, p < 0.002, CI: 0.7-1) and 0.8 (SE: 0.1, p < 0.009, CI: 0.6-1), respectively. Threshold values between 1.7-1.9 for Cho/Cr and 0.8-0.9 for mI/Cr provided high specificity (91% for both metabolites) and adequate sensitivity (75% and 80%, respectively) for discrimination of malignant lesions., Conclusion: A singlet peak at ∼3.8 ppm is present in the majority of tuberculomas and absent in most malignant tumors, potentially a marker to differentiate these lesions. The assignment of the peak is difficult from our analysis; however, guanidinoacetate (Gua) is a possibility. Higher Cho/Cr and mI/Cr ratios should favor malignant lesions over tuberculomas. The presence of lipids and Glx is non-specific., (© The Author(s) 2015.)