Explore the vast range of titles available.

Objectives In the midst of the coronavirus disease 2019 (COVID-19) outbreak, chest X-ray (CXR) imaging is playing an important role in diagnosis and monitoring of patients with COVID-19. We propose a deep learning model for detection of COVID-19 from CXRs, as well as a tool for retrieving similar patients according to the model’s results on their CXRs. For training and evaluating our model, we collected CXRs from inpatients hospitalized in four different hospitals. Methods In this retrospective study, 1384 frontal CXRs, of COVID-19 confirmed patients imaged between March and August 2020, and 1024 matching CXRs of non-COVID patients imaged before the pandemic, were collected and used to build a deep learning classifier for detecting patients positive for COVID-19. The classifier consists of an ensemble of pre-trained deep neural networks (DNNS), specifically, ReNet34, ReNet50¸ ReNet152, and vgg16, and is enhanced by data augmentation and lung segmentation. We further implemented a nearest-neighbors algorithm that uses DNN-based image embeddings to retrieve the images most similar to a given image. Results Our model achieved accuracy of 90.3%, (95% CI: 86.3–93.7%) specificity of 90% (95% CI: 84.3–94%), and sensitivity of 90.5% (95% CI: 85–94%) on a test dataset comprising 15% (350/2326) of the original images. The AUC of the ROC curve is 0.96 (95% CI: 0.93–0.97). Conclusion We provide deep learning models, trained and evaluated on CXRs that can assist medical efforts and reduce medical staff workload in handling COVID-19. Key Points • A machine learning model was able to detect chest X-ray (CXR) images of patients tested positive for COVID-19 with accuracy and detection rate above 90%. • A tool was created for finding existing CXR images with imaging characteristics most similar to a given CXR, according to the model’s image embeddings.

Background: Computed tomography (CT)-derived body composition parameters, including skeletal muscle and fat indices, are prognosticators in oncology. Most studies focus on baseline body-composition parameters; however, changes during treatment may provide better prognostic value. Standardized methods for measuring/reporting these parameters remain limited. Methods: This retrospective study included patients who were treated with immunotherapy for non-small cell lung cancer (NSCLC), renal cell carcinoma (RCC), or melanoma between 2017 and 2024 and had technically adequate baseline and follow-up CT scans. Body composition was analyzed using a novel, fully automated software (CompoCT) for L3 slice selection and segmentation. Body composition indices (e.g., skeletal muscle index [SMI]) were calculated by dividing the cross-sectional area by the patient’s height squared. Results: The cohort included 376 patients (mean [SD] age 66.4 [11.4] years, 67.3% male, 72.6% NSCLC, 14.6% RCC, and 12.8% melanoma). During a median follow-up of 21 months, 220 (58.5%) died. Baseline body composition parameters were not associated with mortality, except for a weak protective effect of higher SMI (HR = 0.98, p = 0.043). In contrast, longitudinal decreases were strongly associated with increased mortality. Relative decreases in SMI (HR, 1.17; 95% CI, 1.07–1.27) or subcutaneous fat index (SFI) (HR, 1.11; 95% CI, 1.07–1.15) significantly increased mortality risk. Multivariate models showed similar concordance (0.65) and identified older age, NSCLC tumor type, and relative decreases in SMI and SFI (per 5% units) as independent predictors of mortality. Conclusions: Longitudinal decreases in skeletal muscle and subcutaneous fat were independent predictors of mortality in immunotherapy-treated patients. Automated CT-based body composition analysis may support treatment decisions during immunotherapy.

Purpose To evaluate the total biopsy and positive biopsy rates in women at high risk of breast cancer compared to the general population. Methods The study group consisted of 330 women with pathogenic variants (PVs) in BRCA1/2 attending the dedicated multidisciplinary breast cancer clinic of a tertiary medical center in Israel. Clinical, genetic, and biopsy data were retrieved from the central healthcare database and the medical files. Patients aged 50 years or older during follow-up were matched 1:10 to women in the general population referred for routine breast cancer screening at the same age, as recommended by international guidelines. The groups were compared for rate of biopsy studies performed and percentage of positive biopsy results. Matched analysis was performed to correct for confounders. Results The total biopsy rate per 1000 follow-up years was 61.7 in the study group and 22.7 in the control group ( p  < 0.001). The corresponding positive biopsy rates per 1000 follow-up years were 26.4 and 2.0 ( p  < 0.001), and the positive biopsy percentages, 42.9% and 8.7% ( p  < 0.0001). Conclusion Women aged 50 + years with PVs in BRCA1/2 attending a dedicated clinic have a 2.7 times higher biopsy rate per 1000 follow-up years, a 13.2 times higher positive biopsy rate per 1000 follow-up years, and a 4.9 times higher positive biopsy percentage than same-aged women in the general population.

Background: Recent advancements in Artificial Intelligence (AI) have the potential to address the challenges of mammographic screening programs by enhancing the performance of Computer-Aided Detection (CAD) systems, improving detection accuracy, and reducing false positive rates and recall rates. These systems were mostly investigated by control trials using cancer-enriched datasets and multiple readers. Objectives: This study aims to evaluate the real-world impact of AI integration on the performance of a breast cancer screening program. Methods: In January 2021, our mammography unit integrated an AI system (iCAD version 2.0) into its mammographic screening protocol. This study evaluates audit data of 31,176 mammograms interpreted between 2017 and 2021, comparing 24,373 mammograms prior to AI implementation and 6803 after the integration. Logistic regression analysis was used to assess the statistical significance of changes in key screening metrics, with a significance level of p < 0.05. Results: This study assesses the impact of artificial intelligence (AI) on mammographic screening. The cancer detection rate increased significantly from 6.2 per 1000 in 2019 to 9.3 per 1000 in 2021, with cancers detected on mammograms rising to 98%. Stage 1 cancer detection reached 100%, and the false negative rate dropped to 0%. Additionally, ductal carcinoma in situ (DCIS) detection decreased from 36.4% in 2019 to 20% in 2021. These findings highlight AI’s effectiveness in improving cancer detection accuracy and efficiency. Conclusions: The integration of AI into mammographic screening demonstrated promising results in improving cancer detection rates and reducing false negative rates. These findings highlight AI’s potential to enhance screening efficacy.

Study objectives: To compare the accuracy of virtual bronchoscopy (VB) with fiberoptic bronchoscopy (FOB) and pulmonary function testing (PFT) for the assessment of tracheal stenosis and bronchial anastomotic stenosis. Design: Prospective case series. Setting: Pulmonary institute of major tertiary university-affiliated center. Patients: The study group included 10 lung transplant recipients and 13 patients with central airway stenosis. Interventions: All patients underwent PFT, VB, and FOB. All cases were graded by each modality on a scale of 1 to 3, and the findings were compared between modalities. Results: Mean ± SD stenosis score was 2.0 ± 0.79 for PFT, 1.62 ± 0.73 for FOB, and 1.82 ± 0.77 for VB. A statistically significant correlation was found between VB and FOB scores (p < 0.0001,r= 0.76) and between VB scores and PFT (p = 0.03,r= 0.45). There was no correlation between PFT and FOB. Conclusions: VB grading of tracheobronchial stenosis is well correlated with PFT. VB may be used to evaluate patients with known tracheobronchial stenosis after treatment and thereby reduce the frequency of repeated invasive FOB performed for that purpose. The correlation of VB with PFT may improve the reliability of this approach.

Objective After lung transplantation, pulmonary artery stenosis (PAS) may occur at the anastomotic site, resulting in poor graft function and hypoxemia. Surgical repair has been the standard-of-care, although percutaneous angioplasty with stent insertion has been performed in patients unsuitable for surgery. We summarize our experience of pulmonary artery stent-graft placement in transplant recipients who were also fit for surgical repair. Materials and Methods Retrospective review of five cases of single-lung transplant recipients (4 male, 1 female, median age 61 years) who underwent percutaneous angioplasty and insertion of stent-graft for severe PAS. Balloon-expandable stent-grafts were used that were tailored to the donor and recipient vessel diameters. Results Stenosis was diagnosed with computed tomography angiography at a median of 44 days (range 22–84) after transplantation. All stent placements were technically successful. There was only one periprocedural complication, a haemothorax that was drained. In four patients, the angioplasty improved the lung function; relative graft perfusion (as assessed by quantitative lung scintigraphy) improved by 26 % (IQR 13–37); and SpO 2 improved by 8 % (IQR 4–9). Conclusion Percutaneous angioplasty using stent-graft is a minimally invasive, safe, and efficacious procedure for treatment of posttransplantation PAS and should be considered as an alternative to surgery even when the patient is considered fit for surgical repair.

Organophosphate poisoning may precipitate complex ventricular arrhythmias, a frequently overlooked and potentially lethal aspect of this condition. Acute effects consist of electrocardiographic ST-T segment changes and AV conduction disturbances of varying degrees, while long-lasting cardiac changes include QT prolongation, polymorphic tachycardia (\"Torsades de Pointes\"), and sudden cardiac death. Cardiac monitoring of organophosphate intoxicated patients for relatively long periods after the poisoning and early aggressive treatment of arrhythmias may be the clue to better survival. We present here a review of the literature with a focus on late cardiac arrhythmias (mainly \"Torsades de pointes\"), possible mechanisms, and treatment modalities, with special emphasis on postpoisoning monitoring for development of arrhythmias.

To compare the accuracy of virtual bronchoscopy (VB) with fiberoptic bronchoscopy (FOB) and pulmonary function testing (PFT) for the assessment of tracheal stenosis and bronchial anastomotic stenosis. Prospective case series. Pulmonary institute of major tertiary university-affiliated center. The study group included 10 lung transplant recipients and 13 patients with central airway stenosis. All patients underwent PFT, VB, and FOB. All cases were graded by each modality on a scale of 1 to 3, and the findings were compared between modalities. Mean +/- SD stenosis score was 2.0 +/- 0.79 for PFT, 1.62 +/- 0.73 for FOB, and 1.82 +/- 0.77 for VB. A statistically significant correlation was found between VB and FOB scores (p < 0.0001, r = 0.76) and between VB scores and PFT (p = 0.03, r = 0.45). There was no correlation between PFT and FOB. VB grading of tracheobronchial stenosis is well correlated with PFT. VB may be used to evaluate patients with known tracheobronchial stenosis after treatment and thereby reduce the frequency of repeated invasive FOB performed for that purpose. The correlation of VB with PFT may improve the reliability of this approach.

Breast cancer is the most common malignancy in women. Mammographic findings such as microcalcifications and masses, as well as morphologic features of masses in sonographic scans, are the main diagnostic targets for tumor detection. However, improved specificity of these imaging modalities is required. A leading alternative target is neoangiogenesis. When pathological, it contributes to the development of numerous types of tumors, and the formation of metastases. Hence, demonstrating neoangiogenesis by visualization of the microvasculature may be of great importance. Super resolution ultrasound localization microscopy enables imaging of the microvasculature at the capillary level. Yet, challenges such as long reconstruction time, dependency on prior knowledge of the system Point Spread Function (PSF), and separability of the Ultrasound Contrast Agents (UCAs), need to be addressed for translation of super-resolution US into the clinic. In this work we use a deep neural network architecture that makes effective use of signal structure to address these challenges. We present in vivo human results of three different breast lesions acquired with a clinical US scanner. By leveraging our trained network, the microvasculature structure is recovered in a short time, without prior PSF knowledge, and without requiring separability of the UCAs. Each of the recoveries exhibits a different structure that corresponds with the known histological structure. This study demonstrates the feasibility of in vivo human super resolution, based on a clinical scanner, to increase US specificity for different breast lesions and promotes the use of US in the diagnosis of breast pathologies.

Early detection of COVID-19 is key in containing the pandemic. Disease detection and evaluation based on imaging is fast and cheap and therefore plays an important role in COVID-19 handling. COVID-19 is easier to detect in chest CT, however, it is expensive, non-portable, and difficult to disinfect, making it unfit as a point-of-care (POC) modality. On the other hand, chest X-ray (CXR) and lung ultrasound (LUS) are widely used, yet, COVID-19 findings in these modalities are not always very clear. Here we train deep neural networks to significantly enhance the capability to detect, grade and monitor COVID-19 patients using CXRs and LUS. Collaborating with several hospitals in Israel we collect a large dataset of CXRs and use this dataset to train a neural network obtaining above 90% detection rate for COVID-19. In addition, in collaboration with ULTRa (Ultrasound Laboratory Trento, Italy) and hospitals in Italy we obtained POC ultrasound data with annotations of the severity of disease and trained a deep network for automatic severity grading.