European Radiology最新文献

Objectives: Prostate cancer (PCa) lesions can be measured on MRI using maximum or biaxial diameters, or as volumes derived by the ellipsoid formula or planimetry. We evaluated the inter- and intra-rater reliability (reliability between different radiologists and the same radiologist during different reading sessions) of lesion size measurements on baseline MRI scans for patients on active surveillance (AS).

Materials and methods: Twenty patients with low- or intermediate-risk PCa (Gleason score 3 + 3 or 3 + 4) and MRI-visible lesions were selected from AS cohorts at two centres (United Kingdom and Italy). Five radiologists, blinded to clinical outcomes and reports, independently measured the index lesion on a baseline MRI scan in a single reading session using: (1) maximum diameter; (2) biaxial diameters; (3) ellipsoid volume, and (4) planimetry volume. Measurements were repeated after a 4-week washout period. Strip plots present lesion size measurements for all methods and readers. Bland-Altman plots were used to present intra-rater reliability.

Results: Graphical presentation of measurements across the twenty patients enabled examination of variability between methods, readers, and reads. There was considerable variation for all methods, and for a single lesion, size measurements spanned previously accepted definitions of clinically significant and insignificant disease. Inter-rater reliability decreased for larger lesions, with notable radiologist-specific differences, and intra-rater reliability appeared better overall.

Conclusion: This study underscores the difficulty of reliably measuring PCa lesions during AS. Intra-rater reliability appeared greater than inter-rater reliability, emphasising that radiologists should remeasure lesions when tracking changes. More work is needed on measuring change in lesion size across serial MRI scans.

Key points: Question Several methods exist for measuring prostate cancer lesion size on MRI for patients on active surveillance, but it is unclear how reliable these methods are. Findings Lesion size measurements varied widely across methods and readers, often spanning thresholds for clinically significant disease, and intra-rater reliability was generally better than inter-rater reliability. Clinical relevance Variability in lesion size measurements on MRI may lead to inconsistent clinical decisions during active surveillance. Our findings emphasise that regardless of the method used, lesions should be remeasured by the same radiologist when monitoring patients on active surveillance.

Objective: This systematic review aims to assess whether studies that followed the 2016 and updated 2024 European Association of Urology (EAU) Renal Cell Carcinoma (RCC) guidelines for CT during follow-up after tumor ablation (TA) yield different oncological outcomes compared to studies that performed more frequent imaging.

Materials and methods: A literature search of relevant search engines was performed up to June 6th, 2025. Studies that reported follow-up schedules of patients after TA for cT1 RCC were included. Studies utilizing more CT scans than recommended by the 2016 and 2024 EAU guidelines were compared with those adhering to the guidelines. Data on recurrences and survival were analyzed.

Results: Thirty-seven studies met the inclusion criteria, involving patients with cT1 RCC treated with TA. The mean 5-year overall survival rate was 82.9%. The pooled recurrence rate was 7.7% in studies that performed more imaging than recommended by the 2016 EAU guideline, compared with 12.3% in studies that adhered to the guideline (p = 0.19). All studies performed more imaging than recommended by the updated 2024 guidelines. Risk of bias was moderate to high in most studies.

Conclusion: The majority of included studies conducted more frequent imaging than advised by the 2016 EAU guidelines, with all studies exceeding the 2024 EAU guidelines. The studies included in our systematic review revealed similar oncological outcomes after TA, among studies that followed the 2016 EAU guidelines and those that performed more frequent imaging, suggesting that more frequent imaging than the 2016 EAU guidelines may not lead to a survival benefit.

Key points: Question Does more frequent follow-up CT imaging after tumor ablation for localized renal cell carcinoma improve oncological outcomes compared to European Association of Urology guideline recommendations? Findings 89% of studies performed more frequent CT scans than the 2016 guidelines; recurrence was 7.7% with extra scans versus 12.3% with guideline adherence. Clinical relevance Current intensive imaging protocols may not improve patient outcomes, supporting potential reduction in follow-up imaging frequency to minimize radiation exposure and healthcare costs while maintaining adequate oncological surveillance.

Objectives: Idiopathic intracranial hypertension (IIH) primarily affects obese women of reproductive age. However, IIH can also occur in individuals outside this typical demographic, where it is associated with a more severe clinical course and poorer visual outcome. Characteristic features of IIH have been identified on cerebral MRI but have not been systematically studied in atypical patient subgroups.

Materials and methods: This retrospective cohort study investigated the prevalence of MRI features of IIH across the following subgroups: males, individuals with normal BMI (< 26 kg/m²), and patients diagnosed above the age of 45. The presence of empty sella (ES), posterior globe flattening (PGF), optic nerve sheath distension (ONSD), optic nerve tortuosity (ONT), transverse sinus stenosis (TSS), DWI-hyperintensity of the optic nerve head (ONH) and ONH-contrast enhancement were evaluated on MRI. The relationship between MRI features and the visual prognosis was investigated.

Results: The study included 172 patients. ES was most frequent with 87%, followed by ONSD in 60%, TSS in 46%, ONT in 39% and PGF in 37%. ONH-DWI hyperintensity was present in 35%, and ONH-contrast enhancement in 22%. The prevalence of MRI features did not significantly differ across demographic subgroups and between atypical and typical IIH. Regression models did not indicate associations between MRI features and visual outcomes.

Conclusions: MRI features of IIH were equally prevalent in typical and atypical demographics. These findings suggest a consistent radiological presentation of IIH across demographic profiles, indicating a shared imaging phenotype regardless of atypical clinical characteristics. Limitations related to the retrospective design warrant future prospective studies.

Key points: Question Do MRI features of IIH differ across demographic subgroups (males, non-obese and older patients) and are they associated with visual outcomes? Findings MRI features of IIH were equally prevalent in typical and atypical patients; however, no significant association between imaging findings and visual outcome was observed. Clinical relevance While MRI supports the diagnosis of IIH across demographic subgroups, established imaging features do not seem to aid in identifying patients at risk of visual deterioration.

Objectives: To evaluate the feasibility, safety, and technical performance of robot-assisted CT-guided cryoablation for pulmonary metastases.

Materials and methods: A single-centre IDEAL stage 2a prospective development study of 26 participants (median age 62 years, IQR 47-71; 14 men) who underwent 30 procedures targeting 37 lung metastases using a robotic navigation system. Median tumour diameter was 9.8 mm (IQR 5.1-12.8). All procedures were performed under general anaesthesia with high-frequency jet ventilation. Feasibility, safety, and technical performance (targeting accuracy, manipulations, radiation dose) were recorded.

Results: Robotic guidance was successfully completed without conversion in 35/37 tumours (95%). One major complication occurred (3%, CTCAE grade 3 pneumothorax requiring 4 days of drainage); all others were grade 1-2. Pneumothoraces were managed by observation (n = 7) or prophylactic intraprocedural chest drain insertion (n = 11). No bronchopleural fistulas were observed. Median hospital stay was 1 night (IQR 1-2). A total of 54 cryoprobes were used. Median Euclidean targeting error on first insertion was 6.1 mm (IQR 2.9-9.7) and lateral error 4.2 mm (IQR 2.2-6.5). The median number of manipulations per probe was 1 (IQR 0-2.5), with one-third requiring no adjustment. Once integrated into the workflow, the "chopstick" technique was frequently applied, supporting conformal ablation. Median total procedure time was 66.5 min (IQR 56.6-92.8). Twelve-month local tumour progression-free survival was 97%.

Conclusion: Robot-assisted CT-guided cryoablation of pulmonary metastases was feasible, safe, and accurate, achieving high targeting precision with minimal cryoprobe manipulation. These findings support evaluation in prospective comparative trials.

Key points: Question Robotic-assisted CT-guided cryoablation of lung metastases is feasible and safe, achieving high targeting accuracy and minimal probe manipulation, even in anatomically challenging cases. Findings Robotic trajectory planning supported complex multiprobe configurations. Procedural refinements-including patient positioning, probe selection, and adoption of "chopstick" configurations-were introduced to address bleeding risk and optimise energy delivery. Clinical relevance Robot-assisted navigation is particularly advantageous in cryoablation, enabling minimal manipulations and accurate probe placement despite the often-necessary complex trajectories.

Objectives: Liver surface nodularity (LSN) is a recognized non-invasive biomarker of cirrhosis. This study introduces auto-LSN, an artificial intelligence (AI)-based algorithm for fully automated LSN quantification, assesses its association with fibrosis stage and its non-inferiority in diagnostic performance for advanced chronic liver disease (ACLD) and cirrhosis compared to the FDA-approved, semi-automated liver boundary analysis (LBA) software.

Materials and methods: This retrospective, bicentric study included patients with chronic liver disease risk factors who underwent CT and liver biopsy between April 2014 and March 2020. Fibrosis stages were grouped into F3-F4 (ACLD) vs F0-F2, and F4 (cirrhosis) vs F0-F3 per the METAVIR. LSN was measured with auto-LSN and LBA. Their association with fibrosis grade and diagnostic accuracy for ACLD and cirrhosis were compared using a -0.05 non-inferiority margin. Mann-Whitney-Wilcoxon tests, Spearman correlation, and area under the receiver operating characteristic curve (AUC) were used.

Results: In 127 patients (68 ± 12 years; 97 men), auto-LSN demonstrated a positive correlation with fibrosis stage (ρ = 0.59; 95% CI [0.48, 0.68]), similar to LBA (ρ = 0.44; 95% CI [0.32, 0.55]), both p < 0.001, with differences within the non-inferiority margin ([0.03, 0.26]). Auto-LSN achieved AUCs of 0.79 (95% CI [0.70, 0.87]) for ACLD and 0.84 (95% CI [0.76, 0.91]) for cirrhosis, comparable to LBA's AUCs of 0.73 (95% CI [0.64, 0.82]) and 0.74 (95% CI [0.66, 0.83]), respectively. All differences were within the non-inferiority margin.

Conclusion: Auto-LSN correlates positively with fibrosis stage and provides non-inferior diagnostic performance compared to LBA. Its full automation and accuracy support its potential for opportunistic screening and objective patient monitoring.

Key points: Question LSN is a key radiological feature for non-invasive ACLD diagnosis. However, current LSN quantification software is only semi-automated, thus time-consuming. Findings The fully automated auto-LSN algorithm for LSN quantification achieved statistically non-inferior diagnostic performance compared to existing semi-automated software for the detection of ACLD and cirrhosis. Clinical relevance Auto-LSN, as a fully automated solution, offers a reliable alternative to existing semi-automated software, enabling mass opportunistic screening of the general population-by evaluating all CT scans performed for any indication-and supporting objective follow-up of at-risk patients.

Objective: Despite advances in mammography screening, some cancers remain undetected, prompting the evaluation of artificial intelligence (AI) as an independent third reader to reduce missed cancers.

Materials and methods: In this prospective study, women eligible for the German Mammography Screening were enrolled at six sites belonging to one screening unit between August 2023 and February 2024. Each mammogram underwent double reading and was independently analyzed using Transpara, an AI-based detection software. Cases rated BI-RADS 4 or 5 by any reader or given a risk score of 10 by the software were reviewed in a consensus conference. Endpoints included: primary-cancer detection rate (CDR) and positive predictive values (PPV); secondary-analysis of cancers detected only by the software or missed by it.

Results: 15,356 female participants (mean age 58.6 ± 5.6 years) were included. Overall, 115 breast cancers were detected (CDR triple reading: 0.75%; 95% CI: 0.62%, 0.90%). CDR of double reading and standalone AI was 0.68% (95% CI: 0.56, 0.83%) and 0.66% (95% CI: 0.54, 0.81%). Using Transpara as a third reader increased the detection rate by 9.5% (95% CI: 4.7%, 16.8%) compared to double reading (p = 0.002). The PPV for consensus-conference referrals was 5.1% (95% CI: 4.2%, 6.1%), lower than double reading 7.5%(95% CI: 6.2%, 9.0%; p < 0.001). For recalled cases, the PPV was 13.7%(95% CI: 11.5%, 16.2%) versus 15.2% (95% CI: 12.6%, 18.1%; p < 0.001). All nine invasive cancers detected solely by AI were Luminal-A-like cancers. Among 13 cancers missed by the software, four were triple-negative.

Conclusion: Adding Transpara as an independent third reader improved detection rates, mainly by identifying additional Luminal-A-like cancers, and increased the workload to the consensus conference and the number of recalled cases.

Key points: Question Does the integration of AI software as an independent third reader improve cancer detection rates in mammography screening without increasing false-positive findings and recall rates? Findings AI as an independent third reader increased cancer detection by 9.5%, mainly identifying Luminal-A-like cancers, significantly decreasing the positive predictive values of cases referred to at the consensus conference and increasing the number of recalled cases. Clinical relevance Using AI as an independent third reader enhances mammographic cancer detection by offering radiologists complementary sensitivity, especially for low-risk lesions. However, maintaining human readers is essential, as AI may miss aggressive subtypes like triple-negative breast cancers.

Objectives: Early recognition of individuals at elevated risk for new ipsilateral ischemic lesions (NIILs) after carotid artery stenting (CAS) is vital for planning effective preventive interventions. The aim of this study was to develop a deep learning (DL) radiomics model to predict NIILs post-CAS from dual-energy CT (DECT) images.

Materials and methods: This study retrospectively enrolled patients from three centers. Carotid plaques were delineated on multiparametric DECT images. A combined model integrating clinical-radiological, handcrafted radiomics (HCR), and DL features was constructed using a support vector machine algorithm to predict NIILs. The model's performance was assessed through the area under the receiver operating characteristic curve (AUC). To improve the interpretability of the model, SHapley Additive exPlanations (SHAP) analysis was applied.

Results: This study involved 336 patients divided into the training (n = 135), internal validation (n = 58), and external test (n = 143) cohorts. NIILs were present in 38.5%, 37.9%, and 39.9% of the subjects, respectively. Symptomatic events and plaque ulceration were identified as independent risk factors for NIILs. The combined model incorporating 2 clinical-radiological risk factors, 9 HCR features, and 15 DL features demonstrated satisfactory performance in predicting NIILs, with AUCs of 0.908, 0.842, and 0.856 in the three cohorts, respectively. The predictions of the combined model were explained both locally and globally by SHAP analysis.

Conclusion: The combined model demonstrated high accuracy in identifying patients at elevated risk for NIILs post-CAS and can serve as an interpretable tool for optimizing treatment strategies.

Key points: Question Early prediction of new ipsilateral ischemic lesions (NIILs) after carotid artery stenting (CAS) is crucial for timely interventions, but no effective, interpretable predictive method exists. Findings The combined model incorporating deep learning radiomics features extracted from multiparametric dual-energy CT images and clinical-radiological features demonstrated high accuracy in predicting NIILs after CAS. Clinical relevance The combined model offers an interpretable tool for identifying patients at high risk for NIILs post-CAS, potentially improving personalized treatment strategies and patient outcomes by enabling targeted preventive care.

Objective: Obstructive hydrocephalus is a critical radiographic finding requiring emergent treatment. Its identification on head CT by an AI model could facilitate sooner life-saving interventions, although there are common co-occurring findings, including intracranial hemorrhage, that can confound this interpretation. This external validation assessed the accuracy of an AI model at identifying obstructive hydrocephalus, including in the presence or absence of other findings.

Materials and methods: This retrospective cohort included 177 thin (≤ 1.5 mm) series and 194 thick (> 1.5 and ≤ 5 mm) series from 200 non-contrast head CT cases. These cases were obtained from patients aged ≥ 18 years at 5 hospitals in the United States. Each case was interpreted independently by up to three neuroradiologists. Each series was then interpreted by the AI model.

Results: The AI model performed with an area under the curve of 0.988 (95% confidence interval (CI): 0.971-0.998) on thin series and 0.986 (95% CI: 0.969-0.997) on thick series. These results were broadly maintained in subgroups for the presence or absence of intracranial hemorrhage, parenchymal abnormality, and ventricular drain, and across demographic and scanner manufacturer subgroups.

Conclusions: The AI model accurately identified obstructive hydrocephalus in this dataset. Its performance in subgroup analyses reflected its robustness.

Key points: Question Can an artificial intelligence model accurately identify obstructive hydrocephalus on head computed tomography, including in the presence or absence of common co-occurring imaging findings? Findings This model accurately identified obstructive hydrocephalus on thin and thick series, including in the presence or absence of intracranial hemorrhage, parenchymal abnormality, and ventricular drain. Clinical relevance This model could assist with triaging abnormal cases, enabling earlier identification and management of obstructive hydrocephalus. Its maintained performance with or without co-occurring findings suggests it specifically identifies obstructive hydrocephalus rather than these findings.