European Radiology Experimental最新文献

Background: Neurofibromatosis type 1 (NF1) is a genetic disorder characterised by skin and nervous system anomalies, primarily involving glial cells and nerve tumours. Pain, particularly chronic pain, is a significant but often overlooked symptom in NF1 patients, affecting their health-related quality of life. The dorsal root ganglion (DRG) is essential for pain signal transmission, yet in vivo studies of DRG in NF1 patients are lacking.

Methods: This prospective study included 20 NF1 patients (8 with neuropathic pain) and 28 healthy controls. Magnetic resonance imaging (MRI) scans of lumbosacral DRG (L5 + S1) were performed using a 3-T scanner. Quantitative MRI techniques were applied to assess DRG volume, T2 relaxation time, and proton density (PD). Statistical analyses compared NF1 patients and controls, and NF1 patients with and without pain.

Results: NF1 patients had a significantly larger DRG volume and higher quantitative T2 and PD values compared to controls. Furthermore, DRG PD was significantly higher in NF1 patients with neuropathic pain than in those without pain. Receiver operator characteristic curve analysis identified DRG PD as the best discriminator of pain in NF1 patients, with an area under the curve of 0.84, indicating relevant and useful discriminatory power.

Conclusion: NF1 patients showed objective macrostructural and microstructural DRG injury changes using dedicated DRG MRI, discriminating neuropathic pain status from non-pain status at the disease-symptom group level. These findings highlight the potential of DRG MRI to quantify DRG pathology in vivo and to determine the risk of functional pain status by imaging.

Relevance statement: The identification of structural and microstructural changes of the DRG by quantitative MRI provides a novel in vivo biomarker for understanding neuropathic pain mechanisms, pain risk assessment and treatment monitoring in NF1.

Key points: Dorsal root ganglia (DRG) in NF1 are enlarged by 176.3% in MRI. In quantitative MRI of DRG NF1, T2 relaxation time is increased by 22.9% and PD by 8.4%. DRG PD can distinguish a painful from a non-painful NF1 phenotype.

Background: The chick chorioallantoic membrane (CAM) model has been utilized for radiofrequency ablation and electroporation, but not yet for cryoablation. This study aims to evaluate the feasibility of the CAM model for preclinical cryoablation research.

Methods: Two cryoablation protocols were established for the study: 120 s-freeze-120 s-thaw-120 s freeze (120 s protocol) and 180 s-freeze-120 s-thaw-180 s freeze (180 s protocol). The study was divided into two parts. First, to evaluate embryo survival, fertilized chicken eggs were incubated. On embryonic day (ED) 12, cryoablation on CAM was performed according to the two protocols. During cryoablation, the temperature of the CAM was recorded using a thermal camera. Embryo survival was monitored until ED 14. Second, to evaluate tumor cryoablation, human neuroendocrine tumor cells (BON-1) were xenografted onto the CAM of fertilized chicken eggs at ED 8. Cryoablation of the xenografted tumors was then performed on ED 12 according to the two protocols. Ablation outcomes were evaluated by stereomicroscopic and histological assessments after harvesting on ED 14.

Results: Embryo survival rates were 8/9 in both protocols. A decrease in the peripheral temperature of 4.5 (± 0.9) °C and 6.7 (± 1.0) °C was observed in the 120 s and 180 s protocols, respectively. Complete ablation of CAM-grown tumors was observed in 2/6 (120 s protocol) and 2/5 (180 s protocol) cases, few scattered tumor cells remaining in 2/6 (120 s protocol) and 2/5 (180 s protocol) cases. Residual interconnected tumor cells were visible in 2/6 (120 s protocol) and 1/5 (180 s protocol) cases.

Conclusion: The CAM model is a feasible platform for preclinical cryoablation studies.

Relevance statement: Chorioallantoic membrane model is a suitable platform for preclinical cryoablation research.

Key points: Chick embryos tolerate the temperature drop during cryoablation well with high survival. Effectiveness of cryoablation on xenografted tumors can be histologically evaluated. Cryoablation protocols for xenografted tumors can be further optimized.

Background: Prostate imaging reporting and data system (PI-RADS) v2.1 guidelines for magnetic resonance imaging acquisition define a standard of 0.40 mm × 0.70 mm in-plane resolution (0.280 mm² pixel area), but adherence has been challenging. We questioned if a modification of a PI-RADS-adherent T2-weighted (T2WI) sequence to one having equivalent pixel area could allow reduced acquisition time but provide improved diagnostic quality (DQ).

Methods: An adherent T2WI sequence was modified by reducing the frequency sampling, thereby reducing the signal bandwidth (BW). This was compensated by increasing the phase sampling for an equivalent pixel area (0.50 mm × 0.57 mm = 0.285 mm²). The BW reduction allowed a two-fold reduction in averaging, also enabling reduced acquisition time. The adherent and modified sequences were evaluated in phantoms and 62 consecutive prostate MRI subjects. Images were evaluated individually by four radiologists using a four-point DQ scale and using prostate imaging quality (PI-QUAL)v2. Each reviewer also indicated any sequence preference. The Wilcoxon test was used.

Results: In the phantom, mean signal-to-noise ratios were equivalent for the two sequences; superior frequency resolution for the adherent sequence, and superior phase resolution for the modified sequence were shown. Across 62 participants, the median acquisition time was reduced by 23%, from 3:55 min:s to 3:01 min:s. For all three means of comparison (DQ, PI-QUALv2, reader preference), the modified sequence was significantly superior (p ≤ 0.037).

Conclusion: Modification of the PI-RADS standard (0.40-mm frequency resolution) to an equivalent, more isotropic pixel area (0.28 mm²) reduced acquisition time and improved image quality.

Relevance statement: Generalization of the PI-RADSv.2.1 minimum technical standard for T2WI in-plane resolution to be more isotropic preserves the targeted high resolution, allowing reduced acquisition time, also reducing motion sensitivity, and improving image quality. This approach may also reduce the need for rescanning poor-quality sequences.

Key points: PI-RADSv2.1 suggests a standard T2WI sequence with 0.40 × 0.70 mm² in-plane resolution. A modified PI-RADSv.2.1-adherent T2WI sequence with equivalent but more isotropic pixel area (0.50 × 0.57 mm²) allowed reduced scan times by 23% and significantly improved DQ. Superiority of the modified sequence appears due to reduced motion sensitivity.

Background: Accurate lung volume determination is crucial for reliable dark-field imaging. We compared different approaches for the determination of lung volume in mean dark-field coefficient calculation.

Methods: In this retrospective analysis of data prospectively acquired between October 2018 and October 2020, patients at least 18 years of age who underwent chest computed tomography (CT) were screened for study participation. Inclusion criteria were the ability to consent and to stand upright without help. Exclusion criteria were pregnancy, lung cancer, pleural effusion, atelectasis, air space disease, ground-glass opacities, and pneumothorax. Lung volume was calculated using four methods: conventional radiography (CR) using shape information; a convolutional neural network (CNN) trained for CR; CT-based volume estimation; and results from pulmonary function testing (PFT). Results were compared using a Student t-test and Spearman ρ correlation statistics.

Results: We studied 81 participants (51 men, 30 women), aged 64 ± 12 years (mean ± standard deviation). All lung volumes derived from the various methods were different from each other: CR, 7.27 ± 1.64 L; CNN, 4.91 ± 1.05 L; CT, 5.25 ± 1.36 L; PFT, 6.54 L ± 1.52 L; p < 0.001 for all comparisons. A high positive correlation was found for all combinations (p < 0.001 for all), the highest one being between CT and CR (ρ = 0.88) and the lowest one between PFT and CNN (ρ = 0.78).

Conclusion: Lung volume and therefore mean dark-field coefficient calculation is highly dependent on the method used, taking into consideration different positioning and inhalation depths.

Relevance statement: This study underscores the impact of the method used for lung volume determination. In the context of mean dark-field coefficient calculation, CR-based methods are more desirable because both dark-field images and conventional images are acquired at the same breathing state, and therefore, biases due to differences in inhalation depth are eliminated.

Key points: Lung volume measurements vary significantly between different determination methods. Mean dark-field coefficient calculations require the same method to ensure comparability. Radiography-based methods simplify workflows and minimize biases, making them most suitable.

Background: To demonstrate that 7-T magnetic resonance imaging (MRI) provides a surrogate for histopathology of fresh ex vivo liver tissue, using the case study of liver fibrosis.

Methods: We prospectively enrolled 20 patients undergoing surgical liver resection between November 2021 and April 2023. Each ex vivo fresh liver tissue specimen (~ 1 cm³) was sectioned in half. The first half, stained using Masson's Trichrome and Perls, was assessed by three pathologists using the METAVIR score (reference standard). The second half was imaged with 7-T MRI using a cryoprobe (fat-suppressed T2-weighted turbo/fast spin-echo sequence, spatial resolution 75 × 75 × 200 µm³) and assessed by three radiologists and the same three pathologists, using a newly developed MRI-METAVIR score.

Results: Five patients were excluded from the final analysis (one patient due to poor specimen quality, two due to surgery cancellation, and two previously published used for reader training). Of the remaining 15 patients, 10 (67%) presented with chronic liver diseases and 8/15 (53%) with advanced (F3 or F4) fibrosis. Radiologists achieved 88% sensitivity, 100% specificity, 93% accuracy (95% confidence interval 68-100%) and 0.94 Harrell's c-index (0.86-1.00). Pathologists achieved 88% sensitivity, 86% specificity, 87% accuracy (60-98%) and 0.87 Harrell's c-index (0.74-0.99). There were no statistically significant differences between MRI-based and pathologic reference standard stage (p ≥ 0.655).

Conclusion: With an in-plane spatial resolution of ~ 75 × 75 µm², MRI paralleled low-magnification histology, enabling the assessment of micro-architectural liver changes, and provided a surrogate for histopathology examination of fresh ex vivo liver tissue samples at a microscopic level.

Relevance statement: 7-T MRI provides a surrogate for histopathology visualisation of fresh ex vivo liver tissue, opening new research perspectives for clinical high-field MRI of the liver.

Key points: Using the newly developed MRI-METAVIR score, 7-T MRI data strongly correlated with histopathology, achieving excellent agreement and accuracy. 7-T MRI accurately differentiated advanced from minimal liver fibrosis. 7-T MRI visualises liver micro-architecture, enabling pathology-like, noninvasive three-dimensional imaging.

Background: Pelvimetry is essential in obstetrics for delivery planning. While computed tomography (CT) is the standard, magnetic resonance imaging (MRI) offers a radiation-free alternative with zero echo time (ZTE) and black bone (BB) sequences providing high bone-to-soft tissue contrast within short scan times. This proof-of-concept study evaluates the reliability of these sequences and the agreement with CT for pelvimetry in a predominantly elderly population.

Methods: This retrospective study included 21 female patients who underwent 3-T whole-body MRI including ZTE and BB sequences and ¹⁸fluorodeoxyglucose (¹⁸F-FDG) positron emission tomography (PET)/CT with optimized low-dose whole-body CT. Obstetric conjugate diameter (OCD), interspinous diameter (ISD), and median transverse diameter (MTD) were measured by five radiologists. Intra-reader, inter-reader, and inter-technique agreement were assessed using intraclass correlation coefficient (ICC) and repeatability/reproducibility coefficients.

Results: Intra-reader agreement was good regardless of diameter or reader: all ICC ≥ 0.90, repeatability ranging from ± 0.26 to ± 0.48 cm (CT), ± 0.30 to ± 0.52 cm (BB), and ± 0.29 to ± 0.67 cm (ZTE). The inter-reader agreement was good regardless of sequence: all ICC ≥ 0.88, reproducibility ranging from ± 0.39 to ± 0.42 (OCD), ± 0.26 to ± 0.51 cm (ISD), and ± 0.53 to ± 0.58 cm (MTD). ZTE and BB showed similar agreement with CT: ± 0.57 to ± 0.81 cm when including inter-reader variability; ± 0.34 to ± 0.47 cm for only intra-reader variability.

Conclusion: ZTE and BB sequences provided reliable measurements with good agreement with CT, for obstetric pelvimetry. Further validation in the context of pregnancy is needed.

Relevance statement: MRI-based pseudo-CT sequences are a promising radiation-free alternative to CT for obstetric pelvimetry, offering the prospect of accurate, reliable measurements of pelvic diameters in pregnant women.

Trial registration: The population included female patients with suspected multiple myeloma from a previous prospective oncology trial (ClinicalTrials.gov: NCT05381077).

Key points: This study explores pseudo-CT MRI sequences for radiation-free non-invasive obstetric pelvimetry. Pseudo-CT zero echo time and black bone sequences provide repeatable and reproducible measurements of pelvic diameters. Pseudo-CT MRI sequences show good inter-technique agreement with the reference CT.

Background: Conventional magnetic resonance diffusion-weighted imaging (DWI) and morphological features have limitations in distinguishing benign from metastatic retropharyngeal lymph nodes (RLNs). We aimed to compare the value of continuous-time random walk (CTRW), fractional-order calculus (FROC), stretched-exponential model (SEM), and conventional DWI, in combination with morphological features, for differentiating between the two groups.

Methods: Fifty-nine patients with 68 RLNs (23 benign and 45 metastatic) were enrolled. All patients underwent DWI with 12 b-values. Diffusion data were reconstructed using conventional DWI, SEM, FROC, and CTRW models, yielding nine parameters: apparent diffusion coefficient (ADC), distributed diffusion coefficient (DDC)_SEM, α_SEM, D_FROC, β_FROC, μ_FROC, D_CTRW, α_CTRW, and β_CTRW. Diffusion parameters and morphological features were compared using Mann-Whitney U, independent sample t, or χ² tests. Logistic regression analysis was performed to identify the best diffusion indicator for classification and to develop a multiparameter model combining morphological features. Area under the receiver operating curve (AUC) and DeLong tests were used.

Results: Significant differences in diffusion parameters were found between benign and metastatic RLNs, except for α_CTRW (p ≤ 0.022). Benign RLNs exhibited higher ADC, DDC_SEM, D_FROC, and D_CTRW, while metastatic RLNs had higher α_SEM, β_FROC, μ_FROC, and β_CTRW. Multivariate logistic regression analysis identified β_CTRW as the optimal single diffusion indicator (AUC = 0.913). The combined model of β_CTRW with morphological features further improved diagnostic performance and yielded an AUC of 0.948.

Conclusion: β_CTRW is an effective noninvasive biomarker for distinguishing between benign and metastatic RLNs. Thus, combining β_CTRW with morphological features enhances diagnostic efficiency.

Relevance statement: This study demonstrates that β_CTRW, derived from the continuous-time random walk diffusion model, when integrated with morphological features, offers a reliable, noninvasive diagnostic approach for differentiating between benign and metastatic retropharyngeal lymph nodes.

Key points: Non-Gaussian diffusion metrics outperformed conventional DWI. β_CTRW was the best indicator for distinguishing benign from metastatic lymph nodes. Combining β_CTRW with minimal axial diameter further improved diagnostic efficiency.

Background: ³¹P-magnetic resonance spectroscopy (MRS) saturation transfer (ST) allows for noninvasive investigation of liver energy metabolism by assessing flux rates of adenosine triphosphate (ATP) synthesis. However, this technique has rarely been applied at clinical field strengths because of long examination times and contamination from muscle tissue. Our aim was to establish a new method to robustly assess ATP synthesis using a clinical scanner.

Methods: A prospective single-center study was performed (January 2023-August 2024) within the German Diabetes Study. We established a suitable ³¹P-MRS ST protocol, tested it in vitro and in vivo and assessed its reproducibility. We assessed the hepatic apparent spin-lattice relaxation time of inorganic phosphate ( $T_{1,Pi}^{\text{'}}$ ), equilibrium forward rate constant ( $k_f$ ), and forward ATP synthesis rate ( $F_{ATP}$ ) in nine control volunteers (CON) (six females) and eight patients (five females) with type 1 diabetes (T1D) and compared differences by ANOVA.

Results: Reproducibility assessment in nine CON, aged 27 ± 4 years (mean ± standard deviation), yielded coefficients of variation for repeated measurements of 7.1% and 21.3% for $T_{1,Pi}^{\text{'}}$ and $k_f$ , respectively. Group comparison revealed higher hepatic $k_f$ (0.34 ± 0.03 s^-1 versus 0.16 ± 0.03 s^-1; p = 0.001) and $F_{ATP}$ (35.3 ± 3.5 mM/min versus 16.4 ± 3.5 mM/min; p = 0.002) in CON than in T1D, aged 42 ± 15 years, respectively.

Conclusion: This ³¹P-MRS ST method allowed for robust assessment of hepatic ATP synthesis at clinical field strength and was sensitive enough to detect differences between CON and T1D volunteers.

Relevance statement: Noninvasive methods to investigate hepatic energy metabolism are urgently needed to evaluate liver health while preventing unnecessary biopsies. For broad clinical applicability, the robustness shown by the proposed method at clinical field strength is crucial.

Trial registration: ClinicalTrials.gov: NCT01055093-Prospective study on diabetes mellitus and its complications in newly diagnosed adult patients (GDC), NCT01055093, Registered: 01/22/2010, https://clinicaltrials.gov/study/N

Background: Gunshot deaths due to homicide or military encounters are a major health concern. Noninvasive bullet characterization is of major importance for patients with lodged bullets or in mass disasters with multiple cadavers, which must be prioritized for autopsy. Therefore, the aim of this study was to investigate whether brass and lead bullets can be differentiated using photon-counting CT (PCCT).

Methods: Nine different lead (n = 6) or brass (n = 3) bullets were investigated on a state-of-the-art PCCT using a clinically unavailable research mode. Here, four image sets were reconstructed for different energy thresholds (20, 55, 72, 90 keV). Three circular regions of interest were placed on the 20-keV threshold images by two readers and automatically copied to the three other threshold images. Based on measured HU mean and max values, dual-energy indices (DEI) were calculated for the low/high energy threshold pairs of 20/90, 55/90, and 72/90 keV.

Results: Significant differences of DEIs between lead and brass projectiles were observed for the 20/90 keV DEI for HU mean ± standard deviation values (Qr40 kernel, lead: -0.085 ± 0.021, brass: 0.024 ± 0.048) and HU max values (Qr40 kernel, lead: -0.093 ± 0.011, brass: 0.023 ± 0.057) (p < 0.001 for both). Differences decreased for the 55/90 and 72/90 keV DEIs between the two projectile materials but remained statistically significant.

Conclusion: In this PCCT phantom study, significant differences were observed between lead and brass bullets in the different energy threshold images.

Relevance statement: Photon-counting CT could be a promising tool for bullet identification as significant differences were found in the different energy threshold images for lead and brass bullets, with application in clinical and forensic radiology.

Key points: In emergency settings, noninvasive bullet characterization is of importance for law enforcement. Bullet material characterization can be performed using photon-counting CT. These characteristics can be quantified in the four different energy threshold images.

Background: The growing demand for follow-up imaging highlights the need for tools supporting the assessment of pulmonary nodules over time. We evaluated the performance of an artificial intelligence (AI)-based system for automated nodule matching.

Methods: In this single-center study, patients with nodules and ≤ 2 chest computed tomography (CT) examinations were retrospectively selected. An AI-based algorithm was used for automated nodule detection and matching. The matching rate and the causes for incorrect matching were evaluated for the ten largest lesions (5-30 mm in diameter) registered on baseline CT. The dependence of the matching rate on nodule number and localization was also analyzed.

Results: One hundred patients (46 females), with a median age of 62 years (interquartile range 57-69), and 253 CTs were included. Focusing on the ten largest lesions, 1,141 lesions were identified, of which 36 (3.2%) were other structures incorrectly identified as nodules (false-positives). Of the 1,105 identified nodules, 964 (87.2%) were correctly detected and matched. The matching rate for nodules registered in both baseline and follow-up scans was 97.8%. The matching rate per case ranged 80.0-100.0% (median 90.0%). Correct matching rate decreased in follow-up examinations to over 50 nodules (p = 0.003), with an overrepresentation of missed matching. Matching rates were higher in parenchymal (91.8%), peripheral (84.4%), and juxtavascular (82.4%) nodules than in juxtaphrenic nodules (71.1%) (p < 0.001). Missed matching was overrepresented in juxtavascular, and incorrect assignment in juxtaphrenic nodules.

Conclusion: The correct automated-matching rate of metastatic pulmonary nodules in follow-up examinations was high, but it depends on localization and a number of nodules.

Relevance statement: The algorithm enables precise follow-up matching of pulmonary nodules, potentially providing a solid basis for standardized and accurate evaluations. Understanding the algorithm's strengths and weaknesses based on nodule localization and number enhances the interpretation of AI-based results.

Key points: The AI algorithm achieved a correct nodule matching rate of 87.2% and up to 97.8% when considering nodules detected in both baseline and follow-up scans. Matching accuracy depended on nodule number and localization. This algorithm has the potential to support response evaluation criteria in solid tumor-based evaluations in clinical practice.