Frontiers in radiology最新文献

Introduction: Cardiac amyloidosis (CA) shares similar clinical and imaging characteristics (e.g., hypertrophic phenotype) with aortic stenosis (AS), but its prognosis is generally worse than severe AS alone. Recent studies suggest that the presence of CA is frequent (1 out of 8 patients) in patients with severe AS. The coexistence of the two diseases complicates the prognosis and therapeutic management of both conditions. Thus, there is an urgent need to standardize and optimize the diagnostic process of CA and AS. The aim of this study is to develop a robust and reliable radiomics-based pipeline to differentiate the two pathologies.

Methods: Thirty patients were included in the study, equally divided between CA and AS. For each patient, a cardiac computed tomography (CCT) was analyzed by extracting 107 radiomics features from the LV wall. Feature robustness was evaluated by means of geometrical transformations to the ROIs and intra-class correlation coefficient (ICC) computation. Various correlation thresholds (0.80, 0.85, 0.90, 0.95, 1), feature selection methods [p-value, least absolute shrinkage and selection operator (LASSO), semi-supervised LASSO, principal component analysis (PCA), semi-supervised PCA, sequential forwards selection] and machine learning classifiers (k-nearest neighbors, support vector machine, decision tree, logistic regression and gradient boosting) were assessed using a leave-one-out cross-validation. Data augmentation was performed using the synthetic minority oversampling technique. Finally, explainability analysis was performed by using the SHapley Additive exPlanations (SHAP) method.

Results: Ninety-two radiomic features were selected as robust and used in the further steps. Best performances of classification were obtained using a correlation threshold of 0.95, PCA (keeping 95% of the variance, corresponding to 9 PCs) and support vector machine classifier reaching an accuracy, sensitivity and specificity of 0.93. Four PCs were found to be mainly dependent on textural features, two on first-order statistics and three on shape and size features.

Conclusion: These preliminary results show that radiomics might be used as non-invasive tool able to differentiate CA from AS using clinical routine available images.

Artificial intelligence (AI) has great potential to increase accuracy and efficiency in many aspects of neuroradiology. It provides substantial opportunities for insights into brain pathophysiology, developing models to determine treatment decisions, and improving current prognostication as well as diagnostic algorithms. Concurrently, the autonomous use of AI models introduces ethical challenges regarding the scope of informed consent, risks associated with data privacy and protection, potential database biases, as well as responsibility and liability that might potentially arise. In this manuscript, we will first provide a brief overview of AI methods used in neuroradiology and segue into key methodological and ethical challenges. Specifically, we discuss the ethical principles affected by AI approaches to human neuroscience and provisions that might be imposed in this domain to ensure that the benefits of AI frameworks remain in alignment with ethics in research and healthcare in the future.

Background: precision-medicine quantitative tools for cross-sectional imaging require painstaking labeling of targets that vary considerably in volume, prohibiting scaling of data annotation efforts and supervised training to large datasets for robust and generalizable clinical performance. A straight-forward time-saving strategy involves manual editing of AI-generated labels, which we call AI-collaborative labeling (AICL). Factors affecting the efficacy and utility of such an approach are unknown. Reduction in time effort is not well documented. Further, edited AI labels may be prone to automation bias.

Purpose: In this pilot, using a cohort of CTs with intracavitary hemorrhage, we evaluate both time savings and AICL label quality and propose criteria that must be met for using AICL annotations as a high-throughput, high-quality ground truth.

Methods: 57 CT scans of patients with traumatic intracavitary hemorrhage were included. No participant recruited for this study had previously interpreted the scans. nnU-net models trained on small existing datasets for each feature (hemothorax/hemoperitoneum/pelvic hematoma; n = 77-253) were used in inference. Two common scenarios served as baseline comparison- de novo expert manual labeling, and expert edits of trained staff labels. Parameters included time effort and image quality graded by a blinded independent expert using a 9-point scale. The observer also attempted to discriminate AICL and expert labels in a random subset (n = 18). Data were compared with ANOVA and post-hoc paired signed rank tests with Bonferroni correction.

Results: AICL reduced time effort 2.8-fold compared to staff label editing, and 8.7-fold compared to expert labeling (corrected p < 0.0006). Mean Likert grades for AICL (8.4, SD:0.6) were significantly higher than for expert labels (7.8, SD:0.9) and edited staff labels (7.7, SD:0.8) (corrected p < 0.0006). The independent observer failed to correctly discriminate AI and human labels.

Conclusion: For our use case and annotators, AICL facilitates rapid large-scale curation of high-quality ground truth. The proposed quality control regime can be employed by other investigators prior to embarking on AICL for segmentation tasks in large datasets.

Osteoporotic compression fractures can be treated with vertebral augmentation. Since intraprocedural pain is common during vertebral body endplate manipulation, these procedures are often performed with conscious sedation or general anesthesia. Research has shown that vertebral endplates are innervated by the basivertebral nerve (BVN), which has been successfully targeted via radiofrequency ablation to treat chronic vertebrogenic lower back pain. With this physiology in mind, we evaluated if temporary BVN block would provide sufficient analgesia so that patients could forego sedation during percutaneous vertebral augmentation. Ten patients with single-level vertebral compression fractures were selected. Prior to balloon augmentation, temporary intraosseous BVN block was achieved using 2% lidocaine injection. All ten patients successfully completed their procedure without intraprocedural sedative or narcotic medications, and without significant deviation from baseline vital signs. Temporary BVN block can be used as intraprocedural anesthesia in select patients who may be poor candidates for general anesthesia or conscious sedation.

Background: Chest x-ray (CXR) is widely applied for the detection and diagnosis of children's lung diseases. Lung field segmentation in digital CXR images is a key section of many computer-aided diagnosis systems.

Objective: In this study, we propose a method based on deep learning to improve the lung segmentation quality and accuracy of children's multi-center CXR images.

Methods: The novelty of the proposed method is the combination of merits of TransUNet and ResUNet. The former can provide a self-attention module improving the feature learning ability of the model, while the latter can avoid the problem of network degradation.

Results: Applied on the test set containing multi-center data, our model achieved a Dice score of 0.9822.

Conclusions: This novel lung segmentation method proposed in this work based on TransResUNet is better than other existing medical image segmentation networks.

Introduction: To date, most mammography-related AI models have been trained using either film or digital mammogram datasets with little overlap. We investigated whether or not combining film and digital mammography during training will help or hinder modern models designed for use on digital mammograms.

Methods: To this end, a total of six binary classifiers were trained for comparison. The first three classifiers were trained using images only from Emory Breast Imaging Dataset (EMBED) using ResNet50, ResNet101, and ResNet152 architectures. The next three classifiers were trained using images from EMBED, Curated Breast Imaging Subset of Digital Database for Screening Mammography (CBIS-DDSM), and Digital Database for Screening Mammography (DDSM) datasets. All six models were tested only on digital mammograms from EMBED.

Results: The results showed that performance degradation to the customized ResNet models was statistically significant overall when EMBED dataset was augmented with CBIS-DDSM/DDSM. While the performance degradation was observed in all racial subgroups, some races are subject to more severe performance drop as compared to other races.

Discussion: The degradation may potentially be due to ( 1) a mismatch in features between film-based and digital mammograms ( 2) a mismatch in pathologic and radiological information. In conclusion, use of both film and digital mammography during training may hinder modern models designed for breast cancer screening. Caution is required when combining film-based and digital mammograms or when utilizing pathologic and radiological information simultaneously.

Purpose: The purpose of this study was to investigate the value of a clinical radiomics model based on Positron emission tomography-computed tomography (PET-CT) radiomics features combined with clinical predictors of Lymphovascular invasion (LVI) in predicting preoperative LVI in patients with colorectal cancer (CRC).

Methods: A total of 95 CRC patients who underwent preoperative ¹⁸F-fluorodeoxyglucose (FDG) PET-CT examination were retrospectively enrolled. Univariate and multivariate logistic regression analyses were used to analyse clinical factors and PET metabolic data in the LVI-positive and LVI-negative groups to identify independent predictors of LVI. We constructed four prediction models based on radiomics features and clinical data to predict LVI status. The predictive efficacy of different models was evaluated according to the receiver operating characteristic curve. Then, the nomogram of the best model was constructed, and its performance was evaluated using calibration and clinical decision curves.

Results: Mean standardized uptake value (SUVmean), maximum tumour diameter and lymph node metastasis were independent predictors of LVI in CRC patients (P < 0.05). The clinical radiomics model obtained the best prediction performance, with an Area Under Curve (AUC) of 0.922 (95%CI 0.820-0.977) and 0.918 (95%CI 0.782-0.982) in the training and validation cohorts, respectively. A nomogram based on the clinical radiomics model was constructed, and the calibration curve fitted well (P > 0.05).

Conclusion: The clinical radiomics prediction model constructed in this study has high value in the preoperative individualized prediction of LVI in CRC patients.

Purpose: To describe our experience with the use of a novel iodized Polyvinyl Alcohol Polymer liquid agent (Easyx) in type II endoleak treatment with translumbar approach.

Methods: Our case series is a retrospective review of patients with type II endoleak (T2E) treated with Easyx from December 2017 to December 2020. Indication for treatment was a persistent T2E with an increasing aneurysm sac ≥5 mm on computed tomography angiography (CTA) over a 6-month interval. Technical success was defined as the embolization of the endoleak nidus with reduction or elimination of the T2E on sequent CTA evaluation. Clinical success was defined as an unchanged or decreased aneurysm sac on follow-up CTA. Secondary endpoints included the presence of artifacts in the postprocedural cross-sectional tomographic imaging and post and intraprocedural complications.

Results: Ten patients were included in our retrospective analysis. All T2E were successfully embolized. Clinical success was achieved in 9 out of 10 patients (90%). The mean follow-up was 14 3-20 months. No beam hardening artifact was observed in follow-up CT providing unaltered imaging.

Conclusion: Easyx is a novel liquid embolic agent with lava-like characteristics and unaltered visibility on subsequent CT examinations. In our initial experience, Easyx showed to have all the efficacy requisites to be an embolization agent for type II EL management. Its efficacy, however, should be evaluated in more extensive studies and eventually compared with other agents.

The aim of this systematic review was to evaluate the state of the art of radiomics in testicular imaging by assessing the quality of radiomic workflow using the Radiomics Quality Score (RQS) and the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2). A systematic literature search was performed to find potentially relevant articles on the applications of radiomics in testicular imaging, and 6 final articles were extracted. The mean RQS was 11,33 ± 3,88 resulting in a percentage of 31,48% ± 10,78%. Regarding QUADAS-2 criteria, no relevant biases were found in the included papers in the patient selection, index test, reference standard criteria and flow-and-timing domain. In conclusion, despite the publication of promising studies, radiomic research on testicular imaging is in its very beginning and still hindered by methodological limitations, and the potential applications of radiomics for this field are still largely unexplored.

Background: Cardiac infiltration is the major predictor of poor prognosis in patients with systemic amyloidosis, thus it becomes of great importance to evaluate cardiac involvement.

Purpose: We aimed to evaluate left ventricular myocardial deformation alteration in patients with cardiac amyloidosis (CA) using layer-specific tissue tracking MR.

Material and methods: Thirty-nine patients with CA were enrolled. Thirty-nine normal controls were also recruited. Layer-specific tissue tracking analysis was done based on cine MR images.

Results: Compared with the control group, a significant reduction in LV whole layer strain values (GLS, GCS, and GRS) and layer-specific strain values was found in patients with CA (all P < 0.01). In addition, GRS and GLS, as well as subendocardial and subepicardial GLS, GRS, and GCS, were all diminished in patients with CA and reduced LVEF, when compared to those with preserved or mid-range LVEF (all P < 0.05). GCS showed the largest AUC (0.9952, P = 0.0001) with a sensitivity of 93.1% and specificity of 90% to predict reduced LVEF (<40%). Moreover, GCS was the only independent predictor of LV systolic dysfunction (Odds Ratio: 3.30, 95% CI:1.341-8.12, and P = 0.009).

Conclusion: Layer-specific tissue tracking MR could be a useful method to assess left ventricular myocardial deformation in patients with CA.