European Radiology Experimental最新文献

Background: R2* and quantitative susceptibility mapping (QSM) are regarded as robust techniques for assessing iron content in the brain. While these techniques are established for normal or moderate iron levels, their usability in extreme iron overload, as seen in aceruloplasminemia (ACP), is unclear. We aimed to evaluate various R2* and QSM algorithms in assessing brain iron levels in patients with ACP compared to healthy controls.

Materials and methods: We acquired a three-dimensional multiecho gradient-echo sequence for R2* and QSM in three patients with ACP and three healthy subjects. Six algorithms each for R2* and QSM were compared. QSM was performed with referencing to whole brain, to cerebrospinal fluid and without referencing. R2* and QSM values were assessed in the caudate nucleus, putamen, globus pallidus, and thalamus.

Results: R2* values varied significantly across algorithms, particularly in the putamen (F(5,50) = 16.51, p < 0.001). For QSM, reference region choice (F(5,150) = 264, p < 0.001) and algorithm selection (F(2,9) = 10, p < 0.001) had an impact on susceptibility values. In patients, referencing to whole brain yielded lower susceptibility values than cerebrospinal fluid (median = 0.147 ppm, range = 0.527 ppm versus median = 0.279 ppm, range = 0.593 ppm).

Conclusion: Extreme iron overload amplifies variability in R2* and QSM measurements. QSM referencing is particularly challenging in diffuse whole-brain iron accumulation; thus, analysis with multiple reference regions might mitigate bias. Both algorithm selection and referencing approaches play a pivotal role in determining measurement accuracy and clinical interpretation under extreme brain iron overload.

Relevance statement: As QSM transitions into clinical use, it will encounter cases of extreme iron overload. Our study in patients with aceruloplasminemia revealed that the choice of reference region significantly influences susceptibility values, with variations exceeding algorithm-dependent differences.

Key points: R2* and QSM vary across algorithms in brain tissue with iron overload. Whole-brain referenced QSM leads to lower susceptibility values in aceruloplasminemia patients. QSM, if properly processed, provides reliable maps in iron overload brain regions. In brain regions with extremely high iron content, R2* mapping might fail.

Background: We compared the performance, confidence, and response consistency of five chatbots powered by large language models in solving European Diploma in Radiology (EDiR) text-based multiple-response questions.

Methods: ChatGPT-4o, ChatGPT-4o-mini, Copilot, Gemini, and Claude 3.5 Sonnet were tested using 52 text-based multiple-response questions from two previous EDiR sessions in two iterations. Chatbots were prompted to evaluate each answer as correct or incorrect and grade its confidence level on a scale of 0 (not confident at all) to 10 (most confident). Scores per question were calculated using a weighted formula that accounted for correct and incorrect answers (range 0.0-1.0).

Results: Claude 3.5 Sonnet achieved the highest score per question (0.84 ± 0.26, mean ± standard deviation) compared to ChatGPT-4o (0.76 ± 0.31), ChatGPT-4o-mini (0.64 ± 0.35), Copilot (0.62 ± 0.37), and Gemini (0.54 ± 0.39) (p < 0.001). A self-reported confidence in answering the questions was 9.0 ± 0.9 for Claude 3.5 Sonnet followed by ChatGPT-4o (8.7 ± 1.1), compared to ChatGPT-4o-mini (8.2 ± 1.3), Copilot (8.2 ± 2.2), and Gemini (8.2 ± 1.6, p < 0.001). Claude 3.5 Sonnet demonstrated superior consistency, changing responses in 5.4% of cases between the two iterations, compared to ChatGPT-4o (6.5%), ChatGPT-4o-mini (8.8%), Copilot (13.8%), and Gemini (18.5%). All chatbots outperformed human candidates from previous EDiR sessions, achieving a passing grade from this part of the examination.

Conclusion: Claude 3.5 Sonnet exhibited superior accuracy, confidence, and consistency, with ChatGPT-4o performing nearly as well. The variation in performance among the evaluated models was substantial.

Relevance statement: Variation in performance, consistency, and confidence among chatbots in solving EDiR test-based questions highlights the need for cautious deployment, particularly in high-stakes clinical and educational settings.

Key points: Claude 3.5 Sonnet outperformed other chatbots in accuracy and response consistency. ChatGPT-4o ranked second, showing strong but slightly less reliable performance. All chatbots surpassed EDiR candidates in text-based EDiR questions.

Background: Accurate segmentation of lung cancer lesions in computed tomography (CT) is essential for precise diagnosis, personalized therapy planning, and treatment response assessment. While automatic segmentation of the primary lung lesion has been widely studied, the ability to segment multiple lesions per patient remains underexplored. In this study, we address this gap by introducing a novel, automated approach for multi-instance segmentation of lung cancer lesions, leveraging a heterogeneous cohort with real-world multicenter data.

Materials and methods: We analyzed 1,081 retrospectively collected CT scans with 5,322 annotated lesions (4.92 ± 13.05 lesions per scan). The cohort was stratified into training (n = 868) and testing (n = 213) subsets. We developed an automated three-step pipeline, including thoracic bounding box extraction, multi-instance lesion segmentation, and false positive reduction via a novel multiscale cascade classifier to filter spurious and non-lesion candidates.

Results: On the independent test set, our method achieved a Dice similarity coefficient of 76% for segmentation and a lesion detection sensitivity of 85%. When evaluated on an external dataset of 188 real-world cases, it achieved a Dice similarity coefficient of 73%, and a lesion detection sensitivity of 85%.

Conclusion: Our approach accurately detected and segmented multiple lung cancer lesions per patient on CT scans, demonstrating robustness across an independent test set and an external real-world dataset.

Relevance statement: AI-driven segmentation comprehensively captures lesion burden, enhancing lung cancer assessment and disease monitoring KEY POINTS: Automatic multi-instance lung cancer lesion segmentation is underexplored yet crucial for disease assessment. Developed a deep learning-based segmentation pipeline trained on multi-center real-world data, which reached 85% sensitivity at external validation. Thoracic bounding box and false positive reduction techniques improved the pipeline's segmentation performance.

Background: This experimental study aimed to compare the image quality of maxillofacial and temporal bone imaging using different radiation dose settings on current high-end CT systems: photon-counting detector CT (PCDCT), dual-source energy-integrating detector CT (DECT), and dual-layer spectral detector CT (SDCT).

Materials and methods: CT scans of a cadaveric human specimen were investigated. Temporal bone imaging was performed with the following parameters: 120 kV and A (high-dose): 140-100 mAs; B (middle-dose): 90-60 mAs; C (low-dose): 50-25 mAs; D (ultra-low-dose): 20-10 mAs. Similarly, for maxillofacial CT: 100 kV and A: 100-80 mAs; B: 70-50 mAs; C: 40-25 mAs; D: 20-10 mAs. Region of interest (ROI)-based noise, SNR, and CNR ratios were calculated for objective assessment of image quality. Subjectively, image quality (IQ) of important anatomic landmarks was assessed using a Likert grading scale from 1 (non-diagnostic) to 5 (excellent).

Results: For temporal bone, PCDCT provided excellent-to-good IQ up to low-dose scans for all anatomical landmarks, which was superior to SDCT (excellent-to-sufficient), followed by DECT (good-to-poor): e.g., for C: 4.3 ± 0.5 versus 3.7 ± 0.6 versus 2.9 ± 0.6, p < 0.001. PCDCT had significantly better IQ compared to SDCT in ultra-low-dose settings (D: 3.9 ± 0.4 versus 2.8 ± 0.4, p < 0.001). For maxillofacial CT, no significant differences in IQ were found between all CT systems using high- and middle-dose scans, e.g., B: 3.9 ± 0.5 versus 3.8 ± 0.7 versus 3.8 ± 0.4 (p = 0.81). In low- and ultra-low-dose settings, IQ was similar by PCDCT and SDCT (C: p = 0.17; D: p = 0.99) and superior to that of DECT (C: p < 0.05).

Conclusion: PCDCT offers excellent image quality for temporal bone and maxillofacial CT even at ultra-low doses; results were, in some cases, superior to SDCT and DECT.

Relevance statement: As PCDCT outperformed modern DECT and SDCT in assessment of maxillofacial and temporal bone CT for image quality and radiation dose, our study suggests that the implementation of PCDCT will improve image quality while reducing radiation exposure in general population.

Key points: This work compares the quality of maxillofacial and temporal bone imaging in PCDCT, DECT, and SDCT. Scans of a cadaveric human specimen were investigated using different radiation doses. PCDCT offers excellent image quality for temporal bone and maxillofacial CT. PCDCT, SDCT, and DECT all showed good image quality overall.

Background: Asbestosis, a rare pneumoconiosis marked by diffuse pulmonary fibrosis, arises from prolonged asbestos exposure. Its diagnosis, guided by the Helsinki criteria, relies on exposure history, clinical findings, radiology, and lung function. However, interobserver variability complicates diagnoses and financial compensation. This study prospectively validated the sensitivity of an AI-driven assessment for asbestosis compensation in the Netherlands. Secondary objectives included evaluating specificity, accuracy, predictive values, area under the curve of the receiver operating characteristic (ROC-AUC), area under the precision-recall curve (PR-AUC), and interobserver variability.

Materials and methods: Between September 2020 and July 2022, 92 adult compensation applicants were assessed using both AI models and pulmonologists' reviews based on Dutch Health Council criteria. The AI model assigned an asbestosis probability score: negative (< 35), uncertain (35-66), or positive (≥ 66). Uncertain cases underwent additional reviews for a final determination.

Results: The AI assessment demonstrated sensitivity of 0.86 (95% confidence interval: 0.77-0.95), specificity of 0.85 (0.76-0.97), accuracy of 0.87 (0.79-0.93), ROC-AUC of 0.92 (0.84-0.97), and PR-AUC of 0.95 (0.89-0.99). Despite strong metrics, the sensitivity target of 98% was unmet. Pulmonologist reviews showed moderate to substantial interobserver variability.

Conclusion: The AI-driven approach demonstrated robust accuracy but insufficient sensitivity for validation. Addressing interobserver variability and incorporating objective fibrosis measurements could enhance future reliability in clinical and compensation settings.

Relevance statement: The AI-driven assessment for financial compensation of asbestosis showed adequate accuracy but did not meet the required sensitivity for validation.

Key points: We prospectively assessed the sensitivity of an AI-driven assessment procedure for financial compensation of asbestosis. The AI-driven asbestosis probability score underperformed across all metrics compared to internal testing. The AI-driven assessment procedure achieved a sensitivity of 0.86 (95% confidence interval: 0.77-0.95). It did not meet the predefined sensitivity target.

Background: Reducing contrast agent and radiation exposure is paramount for pediatric patients. Digital variance angiography (DVA) might address this need by increasing the contrast-to-noise ratio (CNR).

Materials and methods: A total of 132 raw iodinated contrast angiograms of 10 children (mean age: 12 years) who had endovascular procedures for arteriovenous malformations were retrospectively processed for DVA analysis. The CNR of the DVA and digital subtraction angiography (DSA) images was calculated. The visual image quality was assessed using a four-point Likert scale. Statistical analyses were based on the Wilcoxon signed-rank test and one-sample t-test.

Results: The CNR was determined and compared for 3,318 regions of interest in 132 image pairs in four anatomical regions (upper limb (UL), lower limb (LL), head and neck (HN), and chest (CH)). DVA outperformed DSA, with a median overall CNR_DVA/CNR_DSA ratio of 2.00 (UL, 1.83; LL, 1.71; HN, 2.06; CH, 2.23; all p < 0.001). The paired Likert scale scores were significantly different from zero in 50% of the comparisons (in all large vessel and small vessel groups, except in the UL region, and the tissue blush group in the LL and HN regions), indicating a superiority of DSA, but the difference was clinically negligible.

Conclusion: Although DVA improved CNR, it did not surpass DSA in subjective image quality, possibly due to motion artifacts and the high baseline quality of DSA images.

Relevance statement: The enhanced CNR seen with DVA indicates a potential quality reserve that could be exploited to safely reduce contrast agent dose and radiation risks in pediatric patients, who are more susceptible to the long-term effects of radiation.

Key points: In previous studies, DVA was superior to DSA due to a higher CNR and better image quality. However, no evidence was available regarding pediatric endovascular procedures. While DVA exhibited a marked advantage in terms of the CNR, it was unable to surpass DSA in terms of visual assessment. The enhanced CNR seen with DVA indicates a potential quality reserve that could be exploited to safely reduce contrast agent dose and radiation risks in pediatric patients.

Background: We aimed to quantify hepatic vessel volumes across chronic liver disease stages and healthy controls using deep learning-based magnetic resonance imaging (MRI) analysis, and assess correlations with biomarkers for liver (dys)function and fibrosis/portal hypertension.

Methods: We assessed retrospectively healthy controls, non-advanced and advanced chronic liver disease (ACLD) patients using a 3D U-Net model for hepatic vessel segmentation on portal venous phase gadoxetic acid-enhanced 3-T MRI. Total (TVVR), hepatic (HVVR), and intrahepatic portal vein-to-volume ratios (PVVR) were compared between groups and correlated with: albumin-bilirubin (ALBI) and "model for end-stage liver disease-sodium" (MELD-Na) score) and fibrosis/portal hypertension (Fibrosis-4 (FIB-4) Score, liver stiffness measurement (LSM), hepatic venous pressure gradient (HVPG), platelet count (PLT), and spleen volume.

Results: We included 197 subjects, aged 54.9 ± 13.8 years (mean ± standard deviation), 111 males (56.3%): 35 healthy controls, 44 non-ACLD, and 118 ACLD patients. TVVR and HVVR were highest in controls (3.9; 2.1), intermediate in non-ACLD (2.8; 1.7), and lowest in ACLD patients (2.3; 1.0) (p ≤ 0.001). PVVR was reduced in both non-ACLD and ACLD patients (both 1.2) compared to controls (1.7) (p ≤ 0.001), but showed no difference between CLD groups (p = 0.999). HVVR significantly correlated indirectly with FIB-4, ALBI, MELD-Na, LSM, and spleen volume (ρ ranging from -0.27 to -0.40), and directly with PLT (ρ = 0.36). TVVR and PVVR showed similar but weaker correlations.

Conclusion: Deep learning-based hepatic vessel volumetry demonstrated differences between healthy liver and chronic liver disease stages and shows correlations with established markers of disease severity.

Relevance statement: Hepatic vessel volumetry demonstrates differences between healthy liver and chronic liver disease stages, potentially serving as a non-invasive imaging biomarker.

Key points: Deep learning-based vessel analysis can provide automated quantification of hepatic vascular changes across healthy liver and chronic liver disease stages. Automated quantification of hepatic vasculature shows significantly reduced hepatic vascular volume in advanced chronic liver disease compared to non-advanced disease and healthy liver. Decreased hepatic vascular volume, particularly in the hepatic venous system, correlates with markers of liver dysfunction, fibrosis, and portal hypertension.

Background: Dual-energy computed tomography (DECT) detects monosodium urate (MSU) deposits in joints. However, the correlation between coronary atherosclerosis phenotypes and MSU-positive lesions in the cardiovascular system remains unclear. We investigated the correlation between coronary MSU-positive plaques on unenhanced DECT with the coronary atherosclerosis profile at coronary CT angiography.

Methods: One hundred fifty rheumatologic patients were prospectively enrolled. Sixty of them underwent unenhanced DECT and 128-row DECT coronary angiography. Analysis included CAD-RADS stenosis severity, high-risk plaque (HRP) phenotypes, and coronary artery calcium (CAC) score.

Results: Of 60 patients, with a mean age of 63.7 years, including 7 females (11.7%), 37 had gout (61.7%), 9 had hyperuricemia (15%), and 14 had other rheumatologic diseases (23.3%). At DECT, 11 (18.3%) had coronary MSU-positive lesions totaling 24 lesions (left anterior descending, 12; right coronary artery, 10; circumflex, 1; left main, 1). HRP phenotypes were identified in 14 of 60 patients (23.3%). The prevalence of HRP was higher in MSU-positive than MSU-negative patients (63.3% versus 14.2%; p = 0.003; odds ratio 9.91; 95% confidence interval [CI]: 2.30-48.41). CAD-RADS and CAC scores correlated with the number of MSU-positive lesions (ρ = 0.412; 95% CI: 0.167-0.609; p < 0.001) and ρ = 0.412; 95% CI: 0.169-0.609; p < 0.001). None of the major cardiovascular risk factors (smoking, hypertension, dyslipidemia, or diabetes) was associated with MSU-positive lesions.

Conclusion: We found an association between coronary MSU-positive lesions and HRP-phenotypes, as well as a correlation with stenosis severity and calcium burden. MSU-positive lesions may serve as an unenhanced DECT-derived biomarker of increased cardiovascular risk.

Relevance statement: The detection of coronary MSU-positive lesions by DECT could indicate an increased likelihood of HRP phenotypes. These findings suggest their potential as imaging biomarkers for cardiovascular risk, using unenhanced spectral DECT scans or photon-counting CT.

Key points: Identifying gout patients with increased cardiovascular risk remains challenging. Coronary MSU-positive lesions detected on unenhanced DECT may be associated with HRP features on coronary computed tomography angiography. MSU-positive lesions could serve as biomarkers for cardiovascular risk in gout patients.

Background: This study investigated the correlation between decreased internal iliac arterial blood pressure (IIABP) and blood perfusion volume within the vesical artery region during double-balloon-occluded arterial infusion chemotherapy (D-BOAI) for invasive bladder cancer, utilizing two-dimensional perfusion angiography (2D-PA).

Materials and methods: Sixteen patients were enrolled in this study. A double-balloon catheter was positioned into the contralateral internal iliac artery via the femoral artery approach. The catheter's side hole, located between the distal and proximal balloons, facilitated angiographic visualization of the contrast medium (CM) flow into the urinary bladder. Hemodynamic analysis of the CM in the pelvic arteries during D-BOAI was conducted using 2D-PA. Regions of interest (ROIs) were delineated at the side hole (A) as the outflow point for CM and in the vesical artery region (B). The ratio of the area under the curve (AUC) of CM at each ROI (C = B/A) was computed. The decrease in IIABP (D) following balloon occlusion was recorded at the catheter side hole. The relationship between C and D was analyzed using Pearson's product-moment correlation coefficient.

Results: A total of 32 sides from 16 patients were analyzed. The mean C value was 0.39, and the mean D value was 55.2 mmHg, while the mean IIABP post-occlusion measured 66.2 mmHg. A significant positive correlation between C and D was identified, with a correlation coefficient of 0.704 (p < 0.001).

Conclusion: The findings demonstrate a significant positive correlation between blood perfusion volume in the vesical artery region and the reduction in IIABP following balloon occlusion.

Relevance statement: Our results suggest that decreased IIABP after balloon occlusion could result in high concentrations of anticancer drugs in the vesical artery region, and favorable local tumor control in bladder cancer.

Key points: D-BOAI chemotherapy can treat invasive bladder cancer without radical cystectomy. IIABP and flow persist to some extent even following double balloon occlusion. 2D-PA allowed quantitative evaluation of vesical arterial perfusion volume in D-BOAI.

Early accurate diagnosis of osseous and soft tissue injuries following acute knee trauma is crucial for guiding clinical management and preventing chronic instability. Radiography is the appropriate first imaging test applied to detect traumatic osseous injuries. CT is indicated based on clinical symptoms and radiographic concordance. In this acute phase, soft tissue injuries are often clinically overlooked due to swelling and restricted motion, which significantly limit comprehensive physical examination. Moreover, both x-ray and conventional CT imaging are insufficient for addressing this issue due to their limited soft tissue contrast resolution. If clinical suspicion of soft tissue injury persists, an MRI will be performed at a later stage. This may lead to undesirable delays in diagnosis and treatment, thereby potentially impacting patient outcomes. Photon-counting detector CT (PCD-CT) offers enhanced, integrated diagnostic possibilities. The use of spectral imaging data, including color-coded virtual non-calcium (VNCa) images, enables the detection of bone marrow edema (BME) and visualization of key stabilizing soft tissue structures, which may assist emergency department clinicians in determining initial treatment, follow-up, and the need for additional imaging. This technical note illustrates the integral use of ultra-high resolution spectral PCD-CT in a case of a knee injury following an alpine skiing accident. RELEVANCE STATEMENT: The integration of photon-counting detector computed tomography with spectral imaging in acute knee trauma enhances visualization of osseous and soft tissue structures, improving diagnostic accuracy. It may optimize early triage and guide initial treatment for soft tissue injuries. KEY POINTS: Photon-counting detector CT (PCD-CT) enables comprehensive fracture, edema, and soft tissue assessment. Case-based notable correlation between injuries suspected on color-coded spectral imaging and MRI. Photon-counting detector CT (PCD-CT) may enhance early clinical decision-making in knee trauma.