European Radiology最新文献

Purpose: To train and validate machine learning-derived clinical decision algorithm (_MLCDA) for the diagnosis of hyperfunctioning parathyroid glands using preoperative variables to facilitate surgical planning.

Methods: This retrospective study included 458 consecutive primary hyperparathyroidism (PHPT) patients who underwent combined 4D-CT and sestamibi SPECT/CT (MIBI) with subsequent parathyroidectomy from February 2013 to September 2016. The study cohort was divided into training (first 400 patients) and validation sets (remaining 58 patients). Sixteen clinical, laboratory, and imaging variables were evaluated. A random forest algorithm selected the best predictor variables and generated a clinical decision algorithm with the highest performance (_MLCDA). The _MLCDA was trained to predict the probability of a hyperfunctioning vs normal gland for each of the four parathyroid glands in a patient. The reference standard was a four-quadrant location on operative reports and pathology. The accuracy of _MLCDA was prospectively validated.

Results: Of 16 variables, the algorithm selected 3 variables for optimal prediction: combined 4D-CT and MIBI using (1) sensitive reading, (2) specific reading, and (3) cross-product of serum calcium and parathyroid hormone levels and outputted an _MLCDA using five probability categories for hyperfunctioning glands. The _MLCDA demonstrated excellent accuracy for correct classification in the training (4D-CT + MIBI: 0.91 [95% CI: 0.89-0.92]) and validation sets (4D-CT + MIBI: 0.90 [95% CI: 0.86-0.94].

Conclusion: Machine learning generated a clinical decision algorithm that accurately diagnosed hyperfunctioning parathyroid glands through classification into probability categories, which can be implemented for improved preoperative planning and convey diagnostic certainty.

Key points: Question Can an _MLCDA use preoperative variables for the diagnosis of hyperfunctioning parathyroid glands to facilitate surgical planning? Findings The developed _MLCDA demonstrated excellent accuracy for correct classification in the training (0.91 [95% CI: 0.89-0.92]) and validation sets (0.90 [95% CI: 0.86-0.94]). Clinical relevance Using standard preoperative variables, an _MLCDA for diagnosing hyperfunctioning parathyroid glands can be implemented to improve preoperative parathyroid localization and included in radiology reports for surgical planning.

Objective: To investigate imaging biomarkers of tumour response by describing changes in imaging and pathology findings after neoadjuvant radiotherapy (nRT) and exploring their correlations.

Materials and methods: Tumour diameter, volume, and tumour-to-muscle signal intensity (SI) ratio were collected before and after radiotherapy in a cohort of 107 patients with intermediate/high-grade STS and were correlated with post-radiotherapy pathology findings (percentage of necrosis, viable cells, and fibrosis) using Spearman Rank test. Pathological complete response (pCR) was defined as no residual viable cells present, whereas the presence of < 10% viable cells was defined as near-complete pathologic response (near-pCR).

Results: Median amount of necrosis, viable cells, and fibrosis after nRT were 10%, 30%, and 25%, respectively. 7% of patients achieved pCR and 22% near-pCR. No changes in tumour volume were found except for subtypes myxoid liposarcoma (mLPS) -Δ54.47%, undifferentiated pleomorphic sarcoma (UPS) +Δ24.22% and dedifferentiated liposarcoma (dLPS) +Δ35.91%. The median change of tumour-to-muscle SI ratio was -19.7% for the entire population, whereas it was -19.55% and -36.26% for UPS and mLPS, respectively. Correlations (positive and negative) were found between change in volume and the presence of necrosis or fibrosis (r_s = 0.44; r_s = -0.44), as well as between tumour-to-muscle SI ratio and viable cells (r_s = 0.33) or fibrosis (r_s = -0.28).

Conclusion: STS displays extensive heterogeneity in response patterns after nRT. In some subgroups, particularly UPS and mLPS, tumour size changes or tumour-to-muscle SI ratio are significantly linked with the percentage of viable cells, fibrosis, or necrosis.

Key points: Question How do primary soft tissue sarcomas (STS) respond to neoadjuvant therapy, and what correlations exist between pathological findings and imaging characteristics in assessing treatment response? Findings mLPS shrank post-nRT; undifferentiated pleomorphic and dLPSs enlarged. Volume increase correlated with higher necrosis and lower fibrosis; tumour-to-muscle intensity ratio correlated with viable cells. Clinical relevance These findings emphasise the extensive heterogeneity in STS response to nRT across different subtypes. Preoperative correlations between tumour volume and SI changes with necrosis, fibrosis, and viable cells can aid in more precise treatment assessment and prognostication.

Objective: Intensive medical management has been recommended for ischemic stroke of intracranial atherosclerosis (ICAS), but 9.4-15% probability of recurrent stroke remains an inevitable reality. The characteristics of high-risk intracranial plaque that contribute to stroke recurrence after intensive therapy are unclear.

Methods: The patients of acute ischemic stroke due to ICAS from two centers were prospectively analyzed, who underwent the 3D high-resolution head and neck vessel wall magnetic resonance imaging (hr-VW-MRI) at baseline and received intensive medical management within 90 days. The morphological features, such as minimal lumen area (MLA), and histogram parameters including entropy were assessed based on hr-VW-MR images. The recurrence of ischemic events after 6 months was defined as hyperintensity on diffusion-weighted images in the ipsilateral vascular territory. Cox regression analysis was used to calculate the hazard ratio (HR) and 95% confidence interval (CI) for recurrent events.

Results: A total of 222 patients (age 59.5 ± 12.1; males 153) were finally included, and 38 had recurrent stroke after 6 months. After adjusting the age and gender, Cox regression demonstrated that smoking (HR = 4.321; 95% CI, 1.838-10.161; p = 0.001), taking exercise (HR = 0.409; 95% CI, 0.198-0.843; p = 0.015), blood pressure management (HR = 0.180; 95% CI, 0.073-0.443; p = 0.001), MLA (HR = 0.771; 95% CI, 0.625-0.951; p = 0.015) and entropy (HR = 0.274; 95% CI, 0.130-0.576; p = 0.001) were significant predictors of recurrent ischemic stroke. However, the area under curve value of MRI parameters was significantly higher than that of traditional clinical factors (0.86 vs 0.79; p = 0.01).

Conclusions: The plaque characteristics based on hr-VW-MRI may provide complementary values over traditional clinical features in predicting ischemic recurrence for ICAS.

Key points: Question The study addresses recurrent ischemic stroke in intracranial atherosclerosis patients, identifying high-risk plaque features that contribute to recurrence despite intensive medical management. Findings Plaque features on high-resolution vessel wall magnetic resonance imaging (hr-VW-MRI), such as minimal lumen area and entropy, improve prediction of stroke recurrence over clinical factors. Clinical relevance This two-center prospective study improves patient care by using hr-VW-MRI and histogram factors like entropy to better predict stroke recurrence, allowing for more personalized treatment strategies and potentially reducing ischemic events in patients with intracranial atherosclerosis.

Objectives: To evaluate the quality of radiomics research in prostate MRI for the evaluation of prostate cancer (PCa) through the assessment of METhodological RadiomICs (METRICS) score, a new scoring tool recently introduced with the goal of fostering further improvement in radiomics and machine learning methodology.

Materials and methods: A literature search was conducted from July 1st, 2019, to November 30th, 2023, to identify original investigations assessing MRI-based radiomics in the setting of PCa. Seven readers with varying expertise underwent a quality assessment using METRICS. Subgroup analyses were performed to assess whether the quality score varied according to papers' categories (diagnosis, staging, prognosis, technical) and quality ratings among these latter.

Results: From a total of 1106 records, 185 manuscripts were available. Overall, the average METRICS total score was 52% ± 16%. ANOVA and chi-square tests revealed no statistically significant differences between subgroups. Items with the lowest positive scores were adherence to guidelines/checklists (4.9%), handling of confounding factors (14.1%), external testing (15.1%), and the availability of data (15.7%), code (4.3%), and models (1.6%). Conversely, most studies clearly defined patient selection criteria (86.5%), employed a high-quality reference standard (89.2%), and utilized a well-described (85.9%) and clinically applicable (87%) imaging protocol as a radiomics data source.

Conclusion: The quality of MRI-based radiomics research for PCa in recent studies demonstrated good homogeneity and overall moderate quality.

Key points: Question To evaluate the quality of MRI-based radiomics research for PCa, assessed through the METRICS score. Findings The average METRICS total score was 52%, reflecting moderate quality in MRI-based radiomics research for PCa, with no statistically significant differences between subgroups. Clinical relevance Enhancing the quality of radiomics research can improve diagnostic accuracy for PCa, leading to better patient outcomes and more informed clinical decision-making.

Background: There is an increasing need for objective treatment monitoring in perianal fistulising Crohn's disease (pfCD). Therefore, the magnetic resonance novel index for fistula imaging in CD (MAGNIFI-CD) index has been designed and internally validated on the ADMIRE-CD trial cohort. The aim of this study was to externally validate the MAGNIFI-CD index to monitor response to medical and surgical treatment regimens in pfCD.

Methods: A retrospective longitudinal cohort was established of consecutive patients with complex pfCD treated with surgical and/or medical therapy and a baseline and follow-up MRI between January 2007 and May 2021. The MAGNIFI-CD index was scored by two independent, abdominal radiologists blinded for time points and clinical outcomes. Responsiveness, reliability, and test accuracy regarding clinically important improvement were assessed. Cut-offs for response and remission were selected classified on fistula drainage assessment and physician global assessment.

Results: A total of 65 patients (51% female, median age 32 years) were included. A clinically relevant responsiveness of the MAGNIFI-CD was shown, with a significant decrease in clinical remitters and responders with a median MAGNIFI-CD of 18.0 [7.5-20.0] to 9.0 [0.8-16.0] (p < 0.001) and non-significant change in non-responders with a median MAGNIFI-CD of 20.0 [12.0-23.0] to 18.0 [13.0-21.0] (p = 0.22). There was an 'almost perfect' interobserver agreement (ICC = 0.87; 95% CI 0.80-0.92) for the MAGNIFI-CD index. An optimal cut-off value was defined as a decrease of 2 points for clinical response, and a MAGNIFI-CD ≤ 6 for remission at follow-up MRI.

Conclusion: The MAGNIFI-CD index is a responsive and reliable MRI scoring instrument for treatment monitoring in perianal fistulising Crohn's disease.

Clinical relevance statement: The MAGNIFI-CD index is a well-structured, responsive scoring instrument to assess fistula severity and activity that allows quantitative detection of changes in therapy response in patients with perianal fistulising Crohn's disease, thereby facilitating endpoints in clinical trials.

Key points: Well-defined cut-offs for response and remission are needed for objective treatment monitoring of perianal fistulising Crohn's disease (pfCD). Cut-off values for remission and for response at 6 months follow-up were defined. Interobserver agreement was good. The MAGNIFI-CD index is responsive and reliable for treatment monitoring and is suitable for use in clinical trials.

Objectives: To create prediction models (PMs) for distinguishing between benign and malignant liver lesions using quantitative data from dual-energy CT (DECT) without contrast agents.

Materials and methods: This retrospective study included patients with liver lesions who underwent DECT, including non-contrast-enhanced scans. Benign lesions included hepatic hemangioma, whereas malignant lesions included hepatocellular carcinoma, metastatic liver cancer, and intrahepatic cholangiocellular carcinoma. Patients were divided into derivation and validation groups. In the derivation group, two radiologists calculated ten multiparametric data using univariate and multivariate logistic regression to generate PMs. In the validation group, two additional radiologists measured the parameters to assess the diagnostic performance of PMs.

Results: The study included 121 consecutive patients (mean age 67.4 ± 13.8 years, 80 males), with 97 in the derivation group (25 benign and 72 malignant) and 24 in the validation group (7 benign and 17 malignant). Oversampling increased the benign lesion sample to 75, equalizing the malignant group for building PMs. All parameters were statistically significant in univariate analysis (all p < 0.05), leading to the creation of five PMs in multivariate analysis. The area under the curve for the five PMs of two observers was as follows: PM1 (slope K, blood) = 0.76, 0.74; PM2 (slope K, fat) = 0.55, 0.51; PM3 (effective-Z difference, blood) = 0.75, 0.72; PM4 (slope K, blood, fat) = 0.82, 0.78; and PM5 (slope K, effective-Z difference, blood) = 0.90, 0.87. PM5 yielded the best diagnostic performance.

Conclusion: Multiparametric non-contrast-enhanced DECT is a highly effective method for distinguishing between liver lesions.

Clinical relevance statement: The utilization of non-contrast-enhanced DECT is extremely useful for distinguishing between benign and malignant liver lesions. This approach enables physicians to plan better treatment strategies, alleviating concerns associated with contrast allergy, contrast-induced nephropathy, radiation exposure, and excessive medical expenses.

Key points: Distinguishing benign from malignant liver lesions with non-contrast-enhanced CT would be desirable. This model, incorporating slope K, effective Z, and blood quantification, distinguished benign from malignant liver lesions. Non-contrast-enhanced DECT has benefits, particularly in patients with an iodine allergy, renal failure, or asthma.

Objectives: Introducing SPINEPS, a deep learning method for semantic and instance segmentation of 14 spinal structures (ten vertebra substructures, intervertebral discs, spinal cord, spinal canal, and sacrum) in whole-body sagittal T2-weighted turbo spin echo images.

Material and methods: This local ethics committee-approved study utilized a public dataset (train/test 179/39 subjects, 137 female), a German National Cohort (NAKO) subset (train/test 1412/65 subjects, mean age 53, 694 female), and an in-house dataset (test 10 subjects, mean age 70, 5 female). SPINEPS is a semantic segmentation model, followed by a sliding window approach utilizing a second model to create instance masks from the semantic ones. Segmentation evaluation metrics included the Dice score and average symmetrical surface distance (ASSD). Statistical significance was assessed using the Wilcoxon signed-rank test.

Results: On the public dataset, SPINEPS outperformed a nnUNet baseline on every structure and metric (e.g., an average over vertebra instances: dice 0.933 vs 0.911, p < 0.001, ASSD 0.21 vs 0.435, p < 0.001). SPINEPS trained on automated annotations of the NAKO achieves an average global Dice score of 0.918 on the combined NAKO and in-house test split. Adding the training data from the public dataset outperforms this (average instance-wise Dice score over the vertebra substructures 0.803 vs 0.778, average global Dice score 0.931 vs 0.918).

Conclusion: SPINEPS offers segmentation of 14 spinal structures in T2w sagittal images. It provides a semantic mask and an instance mask separating the vertebrae and intervertebral discs. This is the first publicly available algorithm to enable this segmentation.

Key points: Question No publicly available automatic approach can yield semantic and instance segmentation masks for the whole spine (including posterior elements) in T2-weighted sagittal TSE images. Findings Segmenting semantically first and then instance-wise outperforms a baseline trained directly on instance segmentation. The developed model produces high-resolution MRI segmentations for the whole spine. Clinical relevance This study introduces an automatic approach to whole spine segmentation, including posterior elements, in arbitrary fields of view T2w sagittal MR images, enabling easy biomarker extraction, automatic localization of pathologies and degenerative diseases, and quantifying analyses as downstream research.

Objective: Dual-energy computed tomography (DECT) could combine the high-resolution bone window images made available by multi-detector CT technology with its capability to identify bone marrow edema (BME) in the spine, for diagnosing spondylodiscitis. Our objective was to compare the diagnostic performance of contrast-enhanced MRI and non-contrast DECT to identify spondylodiscitis of the thoraco-lumbar spine.

Materials and methods: This prospective study included 77 consecutive participants (39 males; mean age of 61) who underwent DECT and MRI (within 7 days) between January 2020 and October 2023. DECT data were post-processed on a dedicated offline workstation (SyngoVia® VB20) by using a three-material decomposition algorithm. Four radiologists, blinded to clinical data, evaluated non-contrast DECT and contrast-enhanced MRI images. The diagnosis of spondylodiscitis was based on vertebral edema, disc edema, endplate erosions, and paraspinal involvement. Diagnostic accuracy values were calculated by using biopsy as a standard of reference. A multi-reader multi-case analysis was performed.

Results: Biopsy revealed a diagnosis of spondylodiscitis in 46 patients (60%). Thoracic and lumbar spondylodiscitis were diagnosed in 37/46 (80%) and 9/46 (20%) patients, respectively. DECT and MRI overall sensitivity, specificity, and AUC were 0.91, 0.89, and 0.90, and 0.94, 0.93, and 0.93, respectively. At lumbar and thoracic levels, the difference between AUC values between DECT and MRI was not significant (p = 0.15). For DECT and MRI, a very good inter-reader agreement was achieved (k = 0.90 and k = 0.97, respectively).

Conclusions: Contrast-enhanced MRI represents the most accurate imaging tool for the diagnosis of spondylodiscitis. However, only a non-significant drop in diagnostic performance was achieved by evaluating non-contrast DECT images.

Key points: Question To compare the diagnostic performance of contrast-enhanced MRI and non-contrast DECT for identifying spondylodiscitis of the thoraco-lumbar spine. Findings MRI was not significantly superior compared to DECT in diagnosing spondylodiscitis, whereas the inter-reader agreement was near perfect for both MRI and DECT. Clinical relevance DECT represents a fast and accurate imaging tool for the demonstration of BME, erosions, and peri-vertebral inflammation in thoraco-lumbar spondylodiscitis.

Objectives: This single-center retrospective study evaluated the long-term (~5 years) prognostic value of dipyridamole stress cardiac magnetic resonance (CMR) in patients with known or suspected coronary artery disease (CAD), assessing the impact of both key phases of the ischemic cascade (perfusion and wall motion).

Material and methods: We considered 322 consecutive patients who underwent dipyridamole stress CMR. Abnormal wall motion at rest and after dipyridamole, perfusion at stress and at rest, and late gadolinium enhancement (LGE) were analyzed. End-points were non-fatal myocardial infarction, unplanned late revascularization (60 days after CMR), and cardiac death.

Results: Forty-four patients were excluded because they underwent early revascularization (within 60 days after stress CMR), leading to a final population of 278 patients (73 females, 62.42 ± 10.50 years). A positive stress CMR was found in 78 (28.1%) patients; 50 patients had a reversible stress perfusion defect in at least one myocardial segment and 28 had a reversible stress perfusion defect plus worsening of stress wall motion in comparison with the rest. During a mean follow-up time of 59.34 ± 31.72 months, 37 (13.3%) cardiac events were recorded: 10 cardiac deaths, one non-fatal myocardial infarction, and 26 late revascularization after unstable angina or myocardial infarction. According to the Cox regression analysis, age, diabetes mellitus, previous revascularization, left ventricular ejection fraction (LVEF), reversible perfusion and perfusion + motion defect, and LGE were significant univariate prognosticators. The presence of associated wall motion abnormality (WMA) did not provide additional prognostic stratification in comparison to the only perfusion defect. In the multivariate Cox regression, the independent predictive factors were diabetes (hazard-ratio-HR = 5.64, p < 0.0001), reversible perfusion defect and reversible perfusion + motion defect vs normal stress CMR (HR = 6.43, p < 0.0001, and HR = 4.57, p = 0.004; respectively), and LVEF (HR = 0.96, p = 0.010).

Conclusion: A positive dipyridamole stress CMR predicted a higher long-term risk of cardiovascular events, but the presence of inducible WMA did not show any additional prognostic value over the reversible perfusion defect.

Key points: Question The long-term incremental prognostic value of inducible wall motion abnormalities by stress cardiac MR in patients with known or suspected CAD requires investigation. Findings The presence of inducible wall motion abnormalities did not offer additional prognostic value in comparison to the only reversible perfusion defect. Clinical relevance Independent from the presence of wall motion abnormalities, more aggressive management may be appropriate in patients with reversible perfusion defects to reduce the long-term risk of cardiovascular events.

Objectives: To assess glymphatic function and white matter integrity in children with autism spectrum disorder (ASD) using multi-parametric MRI, combined with machine learning to evaluate ASD detection performance.

Materials and methods: This retrospective study collected data from 110 children with ASD (80 exploratory, 43 validation) and 68 typically developing children (50 exploratory, 18 validation) from two centers. The automated diffusion tensor imaging along the perivascular space (aDTI-ALPS), fractional anisotropy (FA), cerebrospinal fluid volume, and perivascular space (PVS) volume indices were extracted from DTI, three-dimensional T1-weighted, and T2-weighted images. Intergroup comparisons were conducted using t-tests, Mann-Whitney U-test, and tract-based spatial statistics. Correlation analysis assessed the relationship between glymphatic function, white matter integrity, and clinical scales. Machine learning models based on MRI indices were developed using the AutoGluon framework.

Results: The PVS volume (p < 0.001) was larger, and aDTI-ALPS index (p < 0.001) was lower in children with ASD compared to typically developing children. FA values were reduced in the ASD group and positively correlated with aDTI-ALPS index. The aDTI-ALPS index correlated with ASD severity (r = -0.27, p = 0.02) and developmental delays (r = 0.63, p < 0.001). Mediation analysis indicated the aDTI-ALPS index partially mediated the relationship between white matter integrity and developmental delay. The MRI-based model achieved an area under the curve of 0.84 for ASD diagnosis.

Conclusion: Analyzing glymphatic function and white matter integrity enhances understanding of ASD's neurobiological underpinnings. The multi-parametric MRI, combined with machine learning, can facilitate the early detection of ASD.

Key points: Question How can multi-parametric MRI based on the glymphatic system improve early diagnosis of autism spectrum disorder (ASD) beyond the limitations of current behavioral assessments? Findings Glymphatic dysfunction and disruptions in white matter integrity were associated with clinical symptoms of ASD. Multi-parametric MRI with machine learning can improve early ASD detection. Clinical relevance Multi-parametric MRI, focusing on glymphatic function and white matter integrity, enhances the diagnostic accuracy of ASD by serving as an objective complement to clinical scales.