European Journal of Radiology最新文献

Objective

This study aimed to quantitatively evaluate the inter-platform reproducibility and longitudinal acquisition repeatability of MRI radiomics features in Fluid-Attenuated Inversion Recovery (FLAIR), T2-weighted (T2W), and T1-weighted (T1W) sequences on MR-Linac systems using an American College of Radiology (ACR) phantom.

Materials and Methods

This study used two MR-Linac systems (A and B) in different cancer centers. The ACR phantom was scanned on system A daily for 30 consecutive days to evaluate longitudinal repeatability. Additionally, retest data were collected after repositioning the phantom. Inter-platform reproducibility was assessed by conducting scans under identical conditions using system B. Regions of interest were delineated on the T1W sequence from system A and mapped to other sequences via rigid registration. Intra-observer and inter-observer comparisons were conducted. Repeatability and reproducibility were assessed using the intraclass correlation coefficient (ICC) and coefficient of variation (CV). Robust radiomics features were identified based on ICC>0.9 and CV<10 %.

Results

Analysis showed that a higher proportion of radiomics features derived from longitudinal FLAIR sequence (51.65 %) met robustness criteria compared to T2W (48.35 %) and T1W (43.96 %). Additionally, more inter-platform features from the FLAIR sequence (62.64 %) were robust compared to T2W (42.86 %) and T1W (39.56 %). Test-retest and intra-observer repeatability were excellent across all sequences, with a median ICC of 0.99 and CV<5%. However, inter-observer reproducibility was inferior, especially for the T1W sequence.

Conclusions

Different sequences show variations in repeatability and reproducibility. The FLAIR sequence demonstrated advantages in both longitudinal repeatability and inter-platform reproducibility. Caution is warranted when interpreting data, particularly in longitudinal or multiplatform radiomics studies.

Background

Magnetic Resonance Imaging (MRI) is frequently used in recent studies on knee osteoarthritis (KOA), focusing on developing innovative MRI-based biomarkers to predict KOA outcomes. The growing volume of publications devoted to this subject highlights the need for an up-to-date review.

Methods

In this narrative review, we utilized the PubMed database to identify studies examining MRI-based biomarkers for the prediction of knee osteoarthritis (KOA), focusing on those reporting relevant prediction, not association, metrics. The identified articles were subsequently categorized into three distinct outcomes: Prediction of KOA incidence (KOAi), KOA progression (KOAp) and total knee arthroplasty risk (TKAr). Within each category, results were organized by the nature of biomarker(s) used, as either quantitative, semi-quantitative or compound.

Results

Due to the lack of predictive metrics such as the area under the ROC curve (AUC) scores, sensitivity or specificity, 27 studies were excluded. A final set of 23 studies were deemed eligible for our analysis. The mean AUC scores reported ranged from 0.67 to 0.83 for predicting KOAi, 0.54 to 0.84 for KOAp and 0.55 to 0.94 for TKAr. Excellent predictive performance (AUC>0.8) was observed for the prediction of radiographic KOAi, KOAp and TKAr when using cartilage and meniscal-based measures, osteophyte scores and infrapatellar fat pad texture, and bone marrow lesions, respectively.

Conclusion

The results showed that numerous studies highlighted the importance of MRI-based biomarkers as promising predictors of the three key outcomes. In addition, this narrative review also emphasized the necessity for KOA prediction studies to include adequate reporting of predictive metrics.

Purpose

To explore the potential differences in epicardial adipose tissue (EAT) volume and attenuation measurements between photon-counting detector (PCD) and energy-integrating detector (EID)-CT systems.

Methods

Fifty patients (mean age 69 ± 8 years, 41 male [82 %]) were prospectively enrolled for a research coronary CT angiography (CCTA) on a PCD-CT within 30 days after clinical EID-based CCTA. EID-CT acquisitions were reconstructed using a Bv40 kernel at 0.6 mm slice thickness. The PCD-CT acquisition was reconstructed at a down-sampled resolution (0.6 mm, Bv40; [PCD-DS]) and at ultra-high resolutions (PCD-UHR) with a 0.2 mm slice thickness and Bv40, Bv48, and Bv64 kernels. EAT segmentation was performed semi-automatically at about 1 cm intervals and interpolated to cover the whole epicardium within a threshold of −190 to −30 HU. A subgroup analysis was performed based on quartile groups created from EID-CT data and PCD-UHR_Bv48 data. Differences were measured using repeated-measures ANOVA and the Friedman test. Correlations were tested using Pearson’s and Spearman’s rho, and agreement using Bland-Altman plots.

Results

EAT volumes significantly differed between some reconstructions (e.g. EID-CT: 138 ml [IQR 100, 188]; PCD-DS: 147 ml [110, 206]; P<0.001). Overall, correlations between PCD-UHR and EID-CT EAT volumes were excellent, e.g. PCD-UHR_Bv48: r: 0.976 (95 % CI: 0.958, 0.987); P<0.001; with good agreement (mean bias: −9.5 ml; limits of agreement [LoA]: −40.6, 21.6). On the other hand, correlations regarding EAT attenuation was moderate, e.g. PCD-UHR_BV48: r: 0.655 (95 % CI: 0.461, 0.790); P<0.001; mean bias: 6.5 HU; LoA: −2.0, 15.0.

Conclusion

EAT attenuation and volume measurements demonstrated different absolute values between PCD-UHR, PCD-DS as well as EID-CT reconstructions, but showed similar tendencies on an intra-individual level. New protocols and threshold ranges need to be developed to allow comparison between PCD-CT and EID-CT data.

Background

Metallic artefacts caused by metal implants, are a common problem in computed tomography (CT) imaging, degrading image quality and diagnostic accuracy. With advancements in artificial intelligence, novel deep learning (DL)-based metal artefact reduction (MAR) algorithms are entering clinical practice.

Objective

This systematic review provides an overview of the performance of the current supervised DL-based MAR algorithms for CT, focusing on three different domains: sinogram, image, and dual domain.

Methods

A literature search was conducted in PubMed, EMBASE, Web of Science, and Scopus. Outcomes were assessed using peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM) or any other objective measure comparing MAR performance to uncorrected images.

Results

After screening, fourteen studies were selected that compared DL-based MAR-algorithms with uncorrected images. MAR-algorithms were categorised into the three domains. Thirteen MAR-algorithms showed a higher PSNR and SSIM value compared to the uncorrected images and to non-DL MAR-algorithms. One study showed statistically significant better MAR performance on clinical data compared to the uncorrected images and non-DL MAR-algorithms based on Hounsfield unit calculations.

Conclusion

DL MAR-algorithms show promising results in reducing metal artefacts, but standardised methodologies are needed to evaluate DL-based MAR-algorithms on clinical data to improve comparability between algorithms.

Clinical relevance statement:

Recent studies highlight the effectiveness of supervised Deep Learning-based MAR-algorithms in improving CT image quality by reducing metal artefacts in the sinogram, image and dual domain. A systematic review is needed to provide an overview of newly developed algorithms.

Purpose

To use transabdominal ultrasound (US) to investigate the impact of posture and axial loading on the lumbar intervertebral disc (IVD) dimensions in healthy adults.

Method

For this single-center, prospective cross-sectional study 54 healthy volunteers (mean age 23.76 ± 3, 26 men) underwent transabdominal US. Lumbar IVD dimensions (height, length, width) at the levels L3/4 and L4/5 were assessed in three test conditions: supine, standing, and standing position with additional axial load of 50 % body weight (standing+50%). Success rates for the longitudinal and transverse US acquisitions and IVD dimension measurements were determined. IVD dimensions were compared across test conditions using two-way repeated measures analysis of variance and post-hoc pairwise t-tests with Bonferroni correction.

Results

The success rate for longitudinal and transverse US acquisition was 100 %. The overall success rate for IVD dimension measurement was 96.4 %, it was highest for IVD height (99.2 %) and lowest for IVD length (93.3 %). IVD height at L4/5 decreased significantly from the supine to standing position (p < 0.05) and from the supine to standing+50% position (p < 0.01). IVD width at L3/4 increased significantly from the supine to standing+50% position (p < 0.05). No significant differences were found for IVD length.

Conclusions

Transabdominal US is a feasible tool to investigate IVD dimensions at L3/4 and L4/5 in different postures and with axial loading. Posture and axial loading significantly influence IVD height and width, but not length.

Purpose

To assess the inter-reader and intra-reader agreement of the Prostate imaging quality version 2 (PI-QUAL v.2) for multiparametric magnetic resonance imaging (mpMRI) among radiologists with varying levels of expertise.

Methods

Fifty men underwent 3 T mpMRI scans in a tertiary referral center. Images were anonymized and assessed by six readers of different expertise (2 expert, 2 basic and 2 beginners) in two sessions: first using PI-QUAL v.2, and then using both PI-QUAL v.2 and v.1 after a 2-week interval. PI-QUAL v.2 scores were considered overall and, for comparison with PI-QUAL v.1, dichotomized according to the threshold of acceptable image quality. Gwet AC₁ index was used to calculate the inter-reader and intra-reader agreement of the scores.

Results

The inter-reader agreement for PI-QUAL v.2 scores was overall moderate (Gwet’s AC₁ = 0.55), being higher for expert readers compared to the beginner and basic ones (Gwet’s AC₁ = 0.66 versus 0.45–0-58). Intra-reader agreement varied from moderate to perfect (Gwet’s AC₁ = 0.43–1.00) and improved with increasing levels of expertise. The ratings were more reproducible for DWI and DCE sequences (Gwet’s AC₁ = 0.62–1.00) compared to T2w (Gwet’s AC₁ = 0.24–0.70). The intra-reader agreement between PI-QUAL v.2 and v.1 scores across readings ranged from almost perfect to perfect (Gwet’s AC₁ = 0.96–1.00).

Conclusions

In a tertiary referral center context, PI-QUAL v.2 is a moderately reliable tool for standardizing prostate mpMRI quality evaluations among readers with varying expertise.

Purpose

Diffusion weighted imaging (DWI) has revealed microstructural changes in lower limb nerves in people with diabetic neuropathy. Microstructural changes in upper limb nerves using DWI in people with diabetes have not yet been explored.

Methods

This cross-sectional study aimed to quantify and compare the microstructure of the median and ulnar nerve in people without diabetes (n = 10), people with diabetes without distal symmetrical polyneuropathy (DSPN; n = 10), people with DSPN in the lower limbs only (DSPN _{FEET ONLY}; n = 12), and people with DSPN in the upper and lower limbs (DSPN _{HANDS & FEET}; n = 9). DSPN diagnosis included electrodiagnosis and corneal confocal microscopy. Tensor metrics, such as fractional anisotropy, radial diffusivity and axial diffusivity, and constrained spherical deconvolution metrics, such as dispersion and complexity, were calculated. Linear mixed-models were used to quantify DWI metrics from multiple models in median and ulnar nerves across the groups, and to evaluate potential differences in metrics at the wrist and elbow based on the principle of a distal-to-proximal disease progression.

Results

Tensor metrics revealed microstructural abnormalities in the median and ulnar nerve in people with DSPN _{HANDS & FEET}, and also already in DSPN _{FEET ONLY}. There were significant negative correlations between electrodiagnostic parameters and tensor metrics. A distal-to-proximal pattern was more pronounced in the median nerve. Non-tensor metrics showed early microstructural changes in people with diabetes without DSPN.

Conclusion

Compared to people without diabetes, microstructural changes in upper limb nerves can be identified in people with diabetes with and without DSPN, even before symptoms occur.

Purpose

Accurate measurements of trabecular bone microarchitecture are required for a proper assessment of bone fragility. Photon-counting detector CT (PCD-CT) has different technical properties than conventional CT, resulting in higher resolution and thereby potentially enabling in-vivo measurement of trabecular microarchitecture. The purpose of this study was to quantify trabecular bone microarchitectural parameters with PCD-CT at varying radiation doses and compare this to µCT as gold standard.

Method

Both distal radii, distal tibiae, femoral heads, and two vertebrae were dissected from one human. All specimens were scanned ex-vivo on a PCD-CT system (slice increment 0.1 mm; pixel size 0.1042–0.127 mm) and a µCT system (isotropic voxel size 49–68.4 µm). The radiation doses of the PCD-CT scans were varied from 2.5 to 120 mGy based on the volume CT dose index (CTDI_vol32). For the PCD-CT scans, contrast-to-noise ratio and trabecular sharpness were calculated and compared between radiation doses. µCT and PCD-CT scans were registered. The trabecular bone was then segmented from all PCD-CT and µCT scans and split into cubes with 6-mm edge length. For each cube, bone volume over total volume, trabecular thickness, trabecular number, and trabecular heterogeneity were calculated and compared between corresponding PCD-CT and µCT cubes.

Results

With increasing dose, contrast-to-noise ratio and trabecular sharpness values increased for the PCD-CT images. Already at the lowest dose, high correlations between the trabecular microarchitectural parameters between µCT and PCD-CT were found (R² = 0.55–0.95), which improved with increasing radiation dose (R² = 0.76–0.96 at 20 mGy).

Conclusions

PCD-CT can be used to quantify trabecular bone microarchitecture, with accuracy comparable to µCT and at clinically relevant radiation doses.

Objectives

To assess the prognostic significance of extra-pancreatic organ invasion in patients with resectable pancreatic ductal adenocarcinoma (PDAC) in the pancreas tail.

Materials & Methods

This retrospective study included patients with resectable PDAC in the pancreas tail who received upfront surgery between 2014 and 2020 at a tertiary institution. Preoperative pancreas protocol computed tomography (CT) scans evaluated tumor size, peripancreatic tumor infiltration, suspicious metastatic lymph nodes, and extra-pancreatic organ invasion. The influence of extra-pancreatic organ invasion, detected by CT or postoperative pathology, on pathologic resection margin status was evaluated using logistic regression. The impact on recurrence-free survival (RFS) was analyzed using multivariable Cox proportional hazard models (clinical-CT and clinical-pathologic).

Results

The study included 158 patients (mean age, 65 years ± 8.8 standard deviation; 93 men). Extra-pancreatic organ invasion identified by either CT (p = 0.92) or pathology (p = 0.99) was not associated with a positive resection margin. Neither CT (p = 0.42) nor pathological (p = 0.64) extra-pancreatic organ invasion independently correlated with RFS. Independent predictors for RFS included suspicious metastatic lymph node (hazard ratio [HR], 2.05; 95 % confidence interval [CI], 1.08–3.9; p = 0.03) on CT in the clinical-CT model, pathological T stage (HR, 2.97; 95 % confidence interval [CI], 1.39–6.35; p = 0.005 for T2 and HR, 3.78; 95 % CI, 1.64–8.76; p = 0.002 for T3) and adjuvant therapy (HR, 0.62; 95 % confidence interval [CI], 0.42–0.92; p = 0.02) in the clinical-pathologic model.

Conclusion

Extra-pancreatic organ invasion does not independently influence pathologic resection margin status and RFS in patients with resectable PDAC in the pancreas tail after curative-intent resection; therefore, it should not be considered a high-risk factor.