Faculty Bibliography

PURPOSE/OBJECTIVE:To evaluate a deep learning based image analysis software for the detection and localization of distal radius fractures. METHOD/METHODS:A deep learning system (DLS) was trained on 524 wrist radiographs (166 showing fractures). Performance was tested on internal (100 radiographs, 42 showing fractures) and external test sets (200 radiographs, 100 showing fractures). Single and combined views of the radiographs were shown to DLS and three readers. Readers were asked to indicate fracture location with regions of interest (ROI). The DLS yielded scores (range 0-1) and a heatmap. Detection performance was expressed as AUC, sensitivity and specificity at the optimal threshold and compared to radiologists' performance. Heatmaps were compared to radiologists' ROIs. RESULTS:The DLS showed excellent performance on the internal test set (AUC 0.93 (95% confidence interval (CI) 0.82-0.98) - 0.96 (0.87-1.00), sensitivity 0.81 (0.58-0.95) - 0.90 (0.70-0.99), specificity 0.86 (0.68-0.96) - 1.0 (0.88-1.0)). DLS performance decreased on the external test set (AUC 0.80 (0.71-0.88) - 0.89 (0.81-0.94), sensitivity 0.64 (0.49-0.77) - 0.92 (0.81-0.98), specificity 0.60 (0.45-0.74) - 0.90 (0.78-0.97)). Radiologists' performance was comparable on internal data (sensitivity 0.71 (0.48-0.89) - 0.95 (0.76-1.0), specificity 0.52 (0.32-0.71) - 0.97 (0.82-1.0)) and better on external data (sensitivity 0.88 (0.76-0.96) - 0.98 (0.89-1.0), specificities 0.66 (0.51-0.79) - 1.0 (0.93-1.0), p < 0.05). In over 90%, the areas of peak activation aligned with radiologists' annotations. CONCLUSIONS:The DLS was able to detect and localize wrist fractures with a performance comparable to radiologists, using only a small dataset for training.

OBJECTIVE. Industry relationships drive technologic innovation in interventional radiology and offer opportunities for professional growth. Women are underrepresented in interventional radiology despite the growing recognition of the importance of diversity. This study characterized gender disparities in financial relationships between industry and academic interventional radiologists. MATERIALS AND METHODS. In this retrospective cross-sectional study, U.S. academic interventional radiology physicians and their academic ranks were identified by searching websites of practices with accredited interventional radiology fellowship programs. Publicly available databases were queried to collect each physician's gender, years since medical school graduation, h-index, academic rank, and industry payments in 2018. Wilcoxon and chi-square tests compared payments between genders. A general linear model assessed the impact of academic rank, years since graduation, gender, and h-index on payments. RESULTS. Of 842 academic interventional radiology physicians, 108 (13%) were women. A total $14,206,599.41 was received by 686 doctors (81%); only $147,975.28 (1%) was received by women. A lower percentage of women (74%) than men (83%) received payments (p = 0.04); median total payments were lower for women ($535) than men ($792) (p = 0.01). Academic rank, h-index, years since graduation, and male gender were independent predictors of higher payments. Industry payments supporting technologic advancement were made exclusively to men. CONCLUSION. Female interventional radiology physicians received fewer and lower industry payments, earning 1% of total payments despite constituting 13% of physicians. Gender independently predicted industry payments, regardless of h-index, academic rank, or years since graduation. Gender disparity in interventional radiology physician-industry relationships warrants further investigation and correction.

A first analysis of simultaneous ⁶⁸Ga-prostate-specific membrane antigen (PSMA)-11 PET/MRI showed some improvement in the detection of recurrent disease at low serum prostate specific antigen (PSA) values below 0.5 ng/mL compared with the already high detection rate of ⁶⁸Ga-PSMA-11 PET/CT. We therefore focused on all patients with biochemical recurrence and PSA values no higher than 0.5 ng/mL to assess the detection rate for ⁶⁸Ga-PSMA-11 PET/MRI. Methods: We retrospectively analyzed a cohort of 66 consecutive patients who underwent ⁶⁸Ga-PSMA-11 PET/MRI for biochemical recurrence with a PSA value no higher than 0.5 ng/mL at our institution. Median PSA level was 0.23 ng/mL (range, 0.03-0.5 ng/mL). Detection of PSMA-positive lesions within the prostate fossa, local and distant lymph nodes, bones, or visceral organs was recorded. In addition, all scans with ⁶⁸Ga-PSMA-11 PET/MRI-positive lesions were retrospectively assessed to analyze if lesions were detected inside or outside a standard salvage radiotherapy volume. Results: Overall, in 36 of 66 patients (54.5%) PSMA-positive lesions were detected; in 26 of 40 (65%) patients with a PSA level between 0.2 and 0.5 ng/mL and in 10 of 26 (38.5%) patients with a PSA level less than 0.2 ng/mL. Even at those low PSA values, only 8 of 66 (12.1%) patients had exclusive local recurrence. Lymph nodes were detected in 23 patients and bone metastases in 5 on ⁶⁸Ga-PSMA-11 PET/MRI. In 26 of 66 patients (39.4%), PSMA-positive lesions were located outside a standard salvage radiotherapy volume. Conclusion: Our data confirm that ⁶⁸Ga-PSMA-11 PET/MRI has a high detection rate for recurrent prostate cancer, even at low PSA levels no higher than 0.5 ng/mL. In addition, we show that ⁶⁸Ga-PSMA-11 PET/MRI detected PSMA-positive lesions outside a standard salvage radiotherapy volume in 39.4% of all patients.

In this prospective study, we quantified the fast pseudo-diffusion contamination by blood perfusion or cerebrospinal fluid (CSF) intravoxel incoherent movements on the measurement of the diffusion tensor metrics in healthy brain tissue. Diffusion-weighted imaging (TR/TE = 4100 ms/90 ms; b-values: 0, 5, 10, 20, 35, 55, 80, 110, 150, 200, 300, 500, 750, 1000, 1300 s/mm², 20 diffusion-encoding directions) was performed on a cohort of five healthy volunteers at 3 Tesla. The projections of the diffusion tensor along each diffusion-encoding direction were computed using a two b-value approach (2b), by fitting the signal to a monoexponential curve (mono), and by correcting for fast pseudo-diffusion compartments using the biexponential intravoxel incoherent motion model (IVIM) (bi). Fractional anisotropy (FA) and mean diffusivity (MD) of the diffusion tensor were quantified in regions of interest drawn over white matter areas, gray matter areas, and the ventricles. A significant dependence of the MD from the evaluation method was found in all selected regions. A lower MD was computed when accounting for the fast-diffusion compartments. A larger dependence was found in the nucleus caudatus (bi: median 0.86 10^-3 mm²/s, Δ2b: -11.2%, Δmono: -14.4%; p = 0.007), in the anterior horn (bi: median 2.04 10^-3 mm²/s, Δ2b: -9.4%, Δmono: -11.5%, p = 0.007) and in the posterior horn of the lateral ventricles (bi: median 2.47 10^-3 mm²/s, Δ2b: -5.5%, Δmono: -11.7%; p = 0.007). Also for the FA, the signal modeling affected the computation of the anisotropy metrics. The deviation depended on the evaluated region with significant differences mainly in the nucleus caudatus (bi: median 0.15, Δ2b: +39.3%, Δmono: +14.7%; p = 0.022) and putamen (bi: median 0.19, Δ2b: +3.1%, Δmono: +17.3%; p = 0.015). Fast pseudo-diffusive regimes locally affect diffusion tensor imaging (DTI) metrics in the brain. Here, we propose the use of an IVIM-based method for correction of signal contaminations through CSF or perfusion.

OBJECTIVE:data from cervical cancer patients. MATERIALS AND METHODS:data. RESULTS:data showed that the optimal threshold was 40 s/mm² for patients with squamous cell carcinoma and a subsampled acquisition of six b-values (scan time, 198 seconds) estimated parameters were not significantly different from reference parameters (individual parameter error rates of less than 5%). In patients with adenocarcinoma, the optimal threshold was 100 s/mm², but an optimal subsample could not be identified. Irrespective of the histological subtype, only three b-values were needed for simplified IVIM, but these parameters did not retain their discriminative ability. CONCLUSION:Subsampling of six b-values halved the IVIM scan time without significant losses in accuracy and discriminative ability. Simplified IVIM is possible with only three b-values, at the risk of losing diagnostic information.

BACKGROUND:F-FDG dose for a 10-min time-of-flight (TOF) PET/MR acquisition that would still allow accurate quantification of supraclavicular BAT volume and activity. METHODS:Twenty datasets from 13 volunteers were retrospectively included from a prospective clinical study. PET emission datasets were modified to simulate step-wise reductions of the original 75 MBq injected dose. The resulting PET images were visually and quantitatively assessed and compared to a 4-min reference scan. For the visual assessment, the image quality and artifacts were scored using a 5-point and a 3-point Likert scale. For the quantitative analysis, image noise and artifacts, BAT metabolic activity, BAT metabolic volume (BMV), and total BAT glycolysis (TBG) were investigated. RESULTS:The visual assessment showed still good image quality for the 35%, 30%, and 25% activity reconstructions with no artifacts. Quantitatively, the background noise was similar to the reference for the 35% and 30% activity reconstructions and the artifacts started to increase significantly in the 25% and lower activity reconstructions. There was no significant difference in supraclavicular BAT metabolic activity, BMV, and TBG between the reference and the 35% to 20% activity reconstructions. CONCLUSIONS:F-FDG tracer dose can be reduced to approximately 19 MBq (25%) while maintaining image quality and accurate supraclavicular BAT quantification. This could decrease the effective dose from 1.4 mSv to 0.36 mSv.

OBJECTIVES:Non-Cartesian Spiral readout can be implemented in 3D Time-of-flight (TOF) MR angiography (MRA) with short acquisition times. In this intra-individual comparison study we evaluated the clinical feasibility of Spiral TOF MRA in comparison with compressed sensing accelerated TOF MRA at 1.5T for intracranial vessel imaging as it has yet to be determined. MATERIALS AND METHODS:Forty-four consecutive patients with suspected intracranial vascular disease were imaged with two Spiral 3D TOFs (Spiral, 0.82x0.82x1.2 mm3, 01:32 min; Spiral 0.8, 0.8x0.8x0.8 mm3, 02:12 min) and a Compressed SENSE accelerated 3D TOF (CS 3.5, 0.82x0.82x1.2 mm3, 03:06 min) at 1.5T. Two neuroradiologists assessed qualitative (visualization of central and peripheral vessels) and quantitative image quality (Contrast Ratio, CR) and performed lesion and variation assessment for all three TOFs in each patient. After the rating process, the readers were questioned and representative cases were reinspected in a non-blinded fashion. For statistical analysis, the Friedman and Nemenyi post-hoc test, Kendall W tests, repeated measure ANOVA and weighted Cohen's Kappa tests were used. RESULTS:The Spiral and Spiral 0.8 outperformed the CS 3.5 in terms of peripheral image quality (p<0.001) and performed equally well in terms of central image quality (p>0.05). The readers noted slight differences in the appearance of maximum intensity projection images. A good to high degree of interstudy agreement between the three TOFs was observed for lesion and variation assessment (W = 0.638, p<0.001 -W = 1, p<0.001). CR values did not differ significantly between the three TOFs (p = 0.534). Interreader agreement ranged from good (K = 0.638) to excellent (K = 1). CONCLUSIONS:Compared to the CS 3.5, both the Spiral and Spiral 0.8 exhibited comparable or better image quality and comparable diagnostic performance at much shorter acquisition times.

BACKGROUND:Brown adipose tissue (BAT) is a specialized form of adipose tissue, able to increase energy expenditure by heat generation in response to various stimuli. Recently, its pathological activation has been implicated in the pathogenesis of cancer cachexia. To establish a causal relationship, we retrospectively investigated the longitudinal changes in BAT and cancer in a large FDG-PET/CT cohort. METHODS:We retrospectively analyzed 13 461 FDG-PET/CT examinations of n = 8 409 patients at our institution from the winter months of 2007-2015. We graded the activation strength of BAT based on the anatomical location of the most caudally activated BAT depot into three tiers, and the stage of the cancer into five general grades. We validated the cancer grading by an interreader analysis and correlation with histopathological stage. Ambient temperature data (seven-day average before the examination) was obtained from a meteorological station close to the hospital. Changes of BAT, cancer, body mass index (BMI) and temperature between the different examinations were examined with Spearman's test and a mixed linear model for correlation, and with a causal inference algorithm for causality. RESULTS:We found n = 283 patients with at least two examinations and active BAT in at least one of them. There was no significant interaction between the changes in BAT activation, cancer burden or BMI. Temperature changes exhibited a strong negative correlation with BAT activity (ϱ = -0.57, p<0.00001). These results were confirmed with the mixed linear model. Causal inference revealed a link of Temperature ➜ BAT in all subjects and also of BMI ➜ BAT in subjects who had lost weight and increased cancer burden, but no role of cancer and no causal links of BAT ➜ BMI. CONCLUSIONS:Our data did not confirm the hypothesis that BAT plays a major role in cancer-mediated weight loss. Temperature changes are the main driver of incidental BAT activity on FDG-PET scans.

Quantitative assessment of functional perfusion capacity and vessel architecture is critical when validating biomaterials for regenerative medicine purposes and requires high-tech analytical methods. Here, combining two clinically relevant imaging techniques, (magnetic resonance imaging; MRI and microcomputed tomography; MicroCT) and using the chorioallantoic membrane (CAM) assay, we present and validate a novel functional and morphological three-dimensional (3D) analysis strategy to study neovascularization in biomaterials relevant for bone regeneration. Using our new pump-assisted approach, the two scaffolds, Optimaix (laminar structure mimicking entities of the diaphysis) and DegraPol (highly porous resembling spongy bone), were shown to directly affect the architecture of the ingrowing neovasculature. Perfusion capacity (MRI) and total vessel volume (MicroCT) strongly correlated for both biomaterials, suggesting that our approach allows for a comprehensive evaluation of the vascularization pattern and efficiency of biomaterials. Being compliant with the 3R-principles (replacement, reduction and refinement), the well-established and easy-to-handle CAM model offers many advantages such as low costs, immune-incompetence and short experimental times with high-grade read-outs when compared to conventional animal models. Therefore, combined with our imaging-guided approach it represents a powerful tool to study angiogenesis in biomaterials.

PURPOSE/OBJECTIVE:To evaluate the interreader variability in prostate and seminal vesicle (SV) segmentation on T2w MRI. METHODS:Six readers segmented the peripheral zone (PZ), transitional zone (TZ) and SV slice-wise on axial T2w prostate MRI examinations of n = 80 patients. Twenty different similarity scores, including dice score (DS), Hausdorff distance (HD) and volumetric similarity coefficient (VS), were computed with the VISCERAL EvaluateSegmentation software for all structures combined and separately for the whole gland (WG = PZ + TZ), TZ and SV. Differences between base, midgland and apex were evaluated with DS slice-wise. Descriptive statistics for similarity scores were computed. Wilcoxon testing to evaluate differences of DS, HD and VS was performed. RESULTS:Overall segmentation variability was good with a mean DS of 0.859 (±SD = 0.0542), HD of 36.6 (±34.9 voxels) and VS of 0.926 (±0.065). The WG showed a DS, HD and VS of 0.738 (±0.144), 36.2 (±35.6 vx) and 0.853 (±0.143), respectively. The TZ showed generally lower variability with a DS of 0.738 (±0.144), HD of 24.8 (±16 vx) and VS of 0.908 (±0.126). The lowest variability was found for the SV with DS of 0.884 (±0.0407), HD of 17 (±10.9 vx) and VS of 0.936 (±0.0509). We found a markedly lower DS of the segmentations in the apex (0.85 ± 0.12) compared to the base (0.87 ± 0.10, p < 0.01) and the midgland (0.89 ± 0.10, p < 0.001). CONCLUSIONS:We report baseline values for interreader variability of prostate and SV segmentation on T2w MRI. Variability was highest in the apex, lower in the base, and lowest in the midgland.