Diagnostic and Interventional Radiology最新文献

Renal cell carcinomas present with locally advanced or metastatic disease in 25% of patients. Thermal ablation may be considered in selected patients with single-site or oligometastatic disease in selected patients. We describe single-session transarterial particle embolization with the assistance of a balloon-occlusion catheter and microwave ablation of a large hypervascular adrenal metastasis using cone beam CT and fluoroscopic XperGuide needle guidance.

Artificial intelligence (AI) and machine learning (ML) are increasingly used in radiology research to deal with large and complex imaging data sets. Nowadays, ML tools have become easily accessible to anyone. Such a low threshold to accessibility might lead to inappropriate usage and misinterpretation, without a clear intention. Therefore, ensuring methodological rigor is of paramount importance. Getting closer to the real-world clinical implementation of AI, a basic understanding of the main concepts should be a must for every radiology professional. In this respect, simplified explanations of the key concepts along with pitfalls and recommendations would be helpful for general radiology community to develop and improve their AI mindset. In this work, twenty-two key issues are reviewed within three categories: pre-modeling, modeling, and post-modeling. Firstly, the concept is shortly defined for each issue. Then, related common pitfalls and best practices are provided. Specifically, the issues included in this paper were validity of scientific question, unrepresentative samples, sample size, missing data, quality of reference standard, batch effect, reliability of features, feature scaling, multi-collinearity, class imbalance, data and target leakage, high-dimensional data, optimization, overfitting, generalization, performance metrics, clinical utility, comparison with conventional statistical and clinical methods, interpretability and explainability, randomness, transparent reporting, and sharing data.

PURPOSE Radiomics can be used to determine the prognosis of gastric cancer (GC). The objective of this study was to predict the disease-free survival (DFS) after GC surgery based on computed tomography-enhanced images combined with clinical features. METHODS Clinical, imaging, and pathological data of patients who underwent gastric adenocarcinoma resection from June 2015 to May 2019 were retrospectively analyzed. The primary outcome was DFS. Radiomics features were selected using Least Absolute Shrinkage and Selection Operator algorithm and converted into the Rad-score. A nomogram was constructed based on the Radscore and other clinical factors. The Rad-score and nomogram were validated in the training and validation groups. RESULTS Totally, 179 patients were randomly divided into the training (n=124) and validation (n=55) groups. In the training group, validation group, and overall population, the Rad-score could be divided into categories indicating low, moderate, and high risk of recurrence, metastasis, or death; all risk categories showed a significant difference between the training, validation, and overall population groups (all P < .001). Positive lymph nodes (hazard ratio (HR)=3.07, 95% CI: 1.52-6.23, P=.002), cancer antigen-125 (HR=3.24, 95% CI: 1.54-6.80, P=.002), and the Radscore (HR=0.73, 95% CI: 0.61-0.87, P < .001) were independently associated with DFS. These 3 variables were used to construct a nomogram. In the training group, the areas under the curve at 3 years were 0.758 and 0.776 for the Rad-score and the nomogram, respectively, while they were both 1.000 in the validation group. The net benefit rate was analyzed using a decision curve in the training and validation groups, and the nomogram was superior to the single Rad-score. CONCLUSION Rad-score is an independent factor for DFS after gastrectomy for GC. The nomogram established in this study could be an effective tool for the clinical prediction of DFS after gastrectomy.

PURPOSE This systematic review and meta-analysis of conventional enhanced magnetic resonance imaging (MRI) were conducted to evaluate the diagnostic performance of imaging features of microvascular invasion (MVI) prediction in hepatocellular carcinoma (HCC). METHODS Relevant studies on diagnosing MVI in HCC by MRI were searched in the MEDLINE, PUBMED, EMBASE, Cochrane library, and Web of Science databases. The pooled mean sensitivity and specificity were calculated using a random effects model. The corresponding positive likelihood ratio (PLR), negative likelihood ratio (NLR), and pooled diagnostic odds ratio (DOR) were calculated. The summary receiver operating characteristic (SROC) curve was used to summarize the overall diagnostic accuracy. Diagnostic performance was evaluated by determining the area under the curve (AUC). Regression analysis by subgroup and sensitivity analysis were used to explore potential sources of heterogeneity. RESULTS A total of 19 studies comprising 1920 HCC patients with 2033 tumors were ultimately enrolled. For the signs of the presence of peritumoral enhancement in the arterial phase, peritumoral hypointensity in the hepatobiliary phase, irregular non-smooth margin, and rim-like enhancement in the arterial phase, the pooled sensitivity values, the pooled specificity values, the pooled PLR values, the pooled NLR values, the pooled DOR values, and the values of the AUC of SROC curves were determined. CONCLUSION The conventional MRI features for predicting MVI showed poor diagnostic performance in HCC. Only signs of the presence of peritumoral enhancement in the arterial phase showed a moderate diagnostic accuracy.

PURPOSE The aim of this study was to compare 18F-fluorodeoxyglucose positron emission tomography- computed tomography (18F-FDG PET/CT) scan with computed tomography (CT) scan for detecting recurrence and metastasis in renal cell carcinoma patients. METHODS This retrospective study included patients from October 2013 to April 2017. Contrast-enhanced CT and PET/CT scans were compared and correlated with histopathology or/and follow-up studies. RESULTS Seventy-six patients, 60 males, were included. Lesions included primary renal, recurrent renal fossa lesions, lymph nodes, and distant metastatic lesions. Of 176 malignant lesions, CT detected 157 lesions; of which, 154 were true positive. Twenty-two false-negative lesions showed abnormal FDG uptake. CT scan had positive predictive value (PPV), negative predictive value (NPV), sensitivity, specificity, and accuracy of 98.0%, 37.1%, 87.5%, 81.2%, and 86.9%, respectively. All 176 lesions were PET/CT-positive. PET/CT had PPV, NPV, sensitivity, specificity, and accuracy of 100% each. The specificity and NPV of PET/CT were superior (P < .05). CONCLUSION PET/CT appears more accurate than CT scan for detecting metastasis and recurrence in renal cell carcinoma patients.

PURPOSE lthough clinical ophthalmologic findings have been reported, no study documented magnetic resonance imaging (MRI) findings in mucopolysaccharidosis (MPS) type VI. The aim of this study was to determine the ophthalmologic imaging findings of MPS type VI in the pediatric age group retrospectively. METHODS Brain MRIs of 10 patients with MPS type VI and 49 healthy children were evaluated independently by two pediatric radiologists for the following characteristics: globe volume, ocular wall thickness, and optic nerve sheath diameter for each orbit. The means of the measurement of each group were compared by using an independent t-test. Agreement and bias between reviewers were assessed by intra-class correlation coefficients (ICC). RESULTS A total of 59 children [32 girls (54.23%), 27 boys (45.77%); age range, 4-16 years; mean age, 10.37 ± 3.73 years] were included in the study. Statistical analysis revealed smaller eyeballs and thicker ocular walls of patients with MPS type VI (P < .001 and P < .001, respectively). However, there was no statistically significant difference in terms of optic nerve sheath diameter between the two groups (P=.648). CONCLUSION Patients with MPS type VI displayed reduced globe volumes and increased ocular wall thicknesses compared to the healthy children. Therefore, we recommend that ophthalmologic imaging findings might prove to be an auxiliary tool in the diagnosis of MPS patients.

PURPOSE This study aimed to verify whether the use of the Kaiser score can improve the diagnostic performance in breast magnetic resonance imaging (MRI) for suspicious lesions and avoid further invasive diagnostic approaches. METHODS This retrospective study enrolled 97 patients who underwent breast MRI before undergoing breast biopsy or surgery. Evaluations were conducted on all MRI images individually by 2 radiologists using the Kaiser score. Neither radiologist had the knowledge of the final histopathological diagnosis. The ability of the Kaiser score in diagnosis was established via a receiver performing characteristic (ROC) analysis, which was measured by the area under the ROC curve (AUC). Youden index was used to define the optimal cutoff value. Kaiser score categories were dichotomized into positive (cutoff score > 4) and negative scores (cutoff score ≤ 4). Cohen's kappa coefficient was used to analyze the inter-rater agreement. RESULTS Histopathology revealed 56 malignant and 41 benign lesions. The AUC for all lesions evaluated by the Kaiser score was 0.992 (95% CI: 0.981-1.0) and 0.958 (95% CI: 0.920-0.996) for 2 radiologists, respectively. Inter-rater agreement of the dichotomized Kaiser score was excellent (κ=0.894, P < .001). A total of 20 lesions (33.8%) previously categorized as BI-RADS 4 were reduced to BI-RADS 2/3 (19 benign lesions and 1 malignant lesion). CONCLUSION The Kaiser score is a valuable auxiliary diagnostic tool for improving the diagnostic ability of radiologists, whose experiences in breast MRI are diverse. In some cases, the application of the Kaiser score could possibly avoid unnecessary breast biopsies.

PURPOSE This article will examine the usefulness of diffusion tensor imaging (DTI) and diffusion-weighted imaging (DWI) on the assessment of axillary lymph nodes (ALN) of breast cancer patients. METHODS Axillary lymph nodes in 66 breast cancer patients were examined by DTI and DWI, and the largest lymph node with increased cortical thickness in axilla was selected. Morphological features, apparent diffusion coefficient (ADC), volume anisotropy, and fractional anisotropy values were measured by using a special software. Imaging findings and histopathological results were recorded. RESULTS Metastatic ALN were detected in 43 (65.1%) patients. Cortical thickness of the metastatic ALN was significantly higher than the non-metastatic ALNs (P < .001), and the long-axis-to-shortaxis ratio was significantly lower in metastatic ALNs (P < .001). There was a statistically significant difference between the ALN status and fatty hilum presence (P < .001). Apparent diffusion coefficient values of metastatic ALNs were statistically lower than those of non-metastatic ALNs (P < .001) using a cutoff value of 1.26 × 10-3 mm2 /s for b=500 ADC and 1.21 × 10-3 mm2 /s for b=800 ADC which had 97.7% sensitivity and 91.3% specificity. Fractional anisotropy and volume anisotropy values were significantly different between both groups. A cutoff value of 0.47 for b-500 fractional anisotropy had 83.7% sensitivity, 69.6% specificity 69.6% positive predictive value, and 83.7% negative predictive value. A cutoff value of 0.33 for b=500 volume anisotropy had 76.7% sensitivity, 78.3% specificity, 86.8% positive predictive value, and 64.3% negative predictive value. CONCLUSION Apparent diffusion coefficient value of metastatic ALNs was found to be significantly lower than those of non-metastatic ALN, and DTI metrics of metastatic ALN were found to be significantly higher than those of non-metastatic ALN. Overall, ADC had a better diagnostic performance than morphological features, fractional anisotropy, and volume anisotropy. Diffusion tensor imagingderived diffusion metrics may be used to complement breast magnetic resonance imaging in the future after further standardization of the imaging parameters.

PURPOSE Thoracic endovascular aortic repair (TEVAR) is a safe and effective treatment method for a variety of thoracic aortic pathologies. We aimed to investigate the mortality and complication outcomes and associated factors of TEVAR treatment in Turkey. METHODS In this single-centered retrospective study, patients with thoracic aorta pathologies treated with TEVAR at Gazi University School of Medicine, Department of Radiology, between January 2009 and January 2020 were included. Perioperative, early, and late mortality, complications, and technical success were the outcomes. RESULTS The sample comprised 58 patients with 68 TEVAR interventions. Eleven (16.2%) patients were female, the mean age was 60.1 ± 13.4 years. Emergent TEVAR was required in 20.7% of the patients. The main indications of TEVAR were intact descending aorta aneurysms in 37.9% of the sample, 31.0% Stanford type-B dissection, and 12.1% traumatic transections. The technical success rate of primary and secondary interventions was 98.3% and 100%, respectively. The mortality rate in the first 30 days was 8.6%. Seventeen (29.3%) cases had at least 1 complication related to TEVAR treatment. The most common complication was type-1A endoleak (10.3%). Having acute symptoms, stroke, and acute renal failure were significantly associated with mortality (P=.020, .049, and .009, respectively). CONCLUSION This study reported the outcomes of TEVAR treatment from a tertiary medical center in Turkey over a decade. Patients presenting with acute symptoms and who developed stroke and acute renal failure after the procedure should be carefully followed up as these factors were found to be associated with mortality.

PURPOSE This study aims to investigate the factors that influence total procedure time when performing computed tomography (CT)-guided percutaneous core-needle lung biopsies. METHODS This is a cross-sectional study of 673 patients, who underwent a CT-guided percutaneous coreneedle biopsy at a tertiary care center from March 2014 to August 2016. Data on patient, nodule, and procedural factors and outcomes were collected retrospectively. Univariate linear regression and a multivariate stepwise linear regression were utilized for analysis. RESULTS Factors most strongly associated with prolonged procedure duration include 20-gauge needle use when compared with 18-gauge needle use (estimated difference in time=1.19), collecting additional core biopsies (estimated difference in time=1.10), decubitus nodule side up (DNSU; estimated difference in time=1.42), and supine positioning (estimated difference in time=1.16) relative to decubitus nodule side down positioning, and increased nodule distance from the skin surface (estimated difference in time=1.03). Increased nodule length (estimated difference in time=0.93) was associated with reductions in procedure duration. Prolonged procedure time was associated with an increased rate of pneumothorax (odds ratio (OR)=1.02; P < .0001) and decreased rate of pulmonary hemorrhage (OR=0.97; P < .0001). CONCLUSION The use of 20-gauge biopsy needle, collecting additional core biopsies, DNSU and supine positioning, smaller nodule size, and increasing nodule distance from the skin surface were associated with increased procedure time for CT-guided core needle biopsies of lung nodules. Prolonged procedure time is associated with a higher rate of pneumothorax and a lower rate of pulmonary hemorrhage.