Semi-Supervised Modified-UNet for Lung Infection Image Segmentation

Automatic lung infection segmentation in computed tomography (CT) scans can offer great assistance in radiological diagnosis by improving accuracy and reducing time required for diagnosis. The biggest challenges for deep learning (DL) models in segmenting infection region are the high variances in infection characteristics, fuzzy boundaries between infected and normal tissues, and the troubles in getting large number of annotated data for training. To resolve such issues, we propose a Modified U-Net (Mod-UNet) model with minor architectural changes and significant modifications in the training process of vanilla 2D UNet. As part of these modifications, we updated the loss function, optimization function, and regularization methods, added a learning rate scheduler and applied advanced data augmentation techniques. Segmentation results on two Covid-19 Lung CT segmentation datasets show that the performance of Mod-UNet is considerably better than the baseline U-Net. Furthermore, to mitigate the issue of lack of annotated data, the Mod-UNet is used in a semi-supervised framework (Semi-Mod-UNet) which works on a random sampling approach to progressively enlarge the training dataset from a large pool of unannotated CT slices. Exhaustive experiments on the two Covid-19 CT segmentation datasets and on a real lung CT volume show that the Mod-UNet and Semi-Mod-UNet significantly outperform other state-of-theart approaches in automated lung infection segmentation. IEEE

HarDNet and Dual-Code Attention Mechanism based model for medical images segmentation

During the formation of medical images, they are easily disturbed by factors such as acquisition devices and tissue backgrounds, causing problems such as blurred image backgrounds and difficulty in differentiation. In this paper, we combine the HarDNet module and the multi-coding attention mechanism module to optimize the two stages of encoding and decoding to improve the model segmentation performance. In the encoding stage, the HarDNet module extracts medical image feature information to improve the segmentation network operation speed. In the decoding stage, the multi-coding attention module is used to extract both the position feature information and channel feature information of the image to improve the model segmentation effect. Finally, to improve the segmentation accuracy of small targets, the use of Cross Entropy and Dice combination function is proposed as the loss function of this algorithm. The algorithm has experimented on three different types of medical datasets, Kvasir-SEG, ISIC2018, and COVID-19CT. The values of JS were 0.7189, 0.7702, 0.9895, ACC were 0.8964, 0.9491, 0.9965, SENS were 0.7634, 0.8204, 0.9976, PRE were 0.9214, 0.9504, 0.9931. The experimental results showed that the model proposed in this paper achieved excellent segmentation results in all the above evaluation indexes, which can effectively assist doctors to diagnose related diseases quickly and improve the speed of diagnosis and patients’quality of life. Author

Past, present and future of RTT training in mainland China: Evaluation of trainee perspectives

Objective: The radiation therapy technologist fundamentals training program (RFTP) facilitates knowledge and skills development of newly employed radiation therapy technologists (RTTs) within our China network. Since its initial implementation in 2019, the RFTP has evolved to address the diversity of RTTs' education and experience, as well as changing local clinical contexts. In particular, a shift to remote delivery and assessment has been required during the COVID-19 pandemic. This quality improvement initiative aimed to evaluate the impact of the RFTP on learning engagement and outcomes, from trainee perspectives. Method(s): Online pre-interview surveys and semi-structured interviews were conducted with 16 RTTs from five China sites in July and August 2022.1 Participants provided verbal informed consent regarding the survey and interview recordings for subsequent analysis. Surveys were reported with descriptive statistics, and interview themes were developed using direct content analysis.2 Results: 15/16 participants qualified in a non-RTT field of study, with most practiced in medical imaging (N = 7);12/12 participants with previous RTT experience reported differences in practice standards. All participants rated the RFTP highly (see Figure), with IGRT (13/16), ARIA (14/16), and SimCT (7/16) most frequently identified as new areas of learning;14 participants who completed the RFTP reported the preparation for IGRT standard workflow was most valuable. Discussion/Conclusion: Results show that the RFTP is an essential on-boarding program that advances RTTs' knowledge and reduces the skills gap to perform our network's established workflows. Additional feedback gained through this initiative will be considered for future development of the RFTP.

Shaping radiographers of the future: A snapshot of contemporary medical imaging education

Radiographers function in a niche environment, blending advanced technical skills with patient-focussed care in a multi-disciplinary environment. The past decade has brought significant change to the profession with further change projected into the future. Education programs are dynamic, responsive to emerging technologies and improve from each iteration. The education experience of current students is significantly different to that of experienced radiographers. This presentation provides a snapshot of contemporary education approaches in a medical imaging undergraduate program, preparing radiographers for the future. A significant component of skill development in medical imaging degrees is achieved through clinical placement across a range of settings. Education providers work with clinical departments to maximise learning opportunities, scaffolding structured progression of learning. Lack of availability of suitable resources or placement opportunities, and the priority that must be given to patient service delivery can be challenging for clinical experiences. COVID-19 pandemic restrictions have exacerbated issues, particularly for already time poor clinical environments. The education program showcased draws on contemporary best imaging practice, curriculum design and learning and teaching approaches. For example, the teaching team have applied simulation as a purposeful technique to add safe and procedural steps as part of a cohesive whole-of-curriculum student learning. Simulation prior to clinical placement is widely recognised as an effective response to the challenges of assuring safe and skilled practice.1 In turn, students can reflect on this experiential learning providing robust feedback and discussion using Kolb's reflective practice - exploring the impact of their learning on their future practice.2.

Thyroid Incidentalomas in Hospitalized Patients With COVID-19: A Single-Center Retrospective Analysis.

Context: During the COVID-19 pandemic, both people with underlying diseases and previously healthy people were infected with SARS-CoV-2. In our institute, most hospitalized patients underwent chest computed tomography (CT) to evaluate pulmonary involvement and complication of COVID-19. There are currently limited data regarding thyroid CT incidentalomas in healthy people. Objective: We aimed to investigate the prevalence and predictors of thyroid incidentalomas among hospitalized patients with COVID-19. Methods: A single-center retrospective study included hospitalized patients aged ≥15 years with COVID-19 who underwent chest CT during April 2020 and October 2021. Thyroid incidentalomas were reviewed and identified by an experienced radiologist. Logistic regression analysis was used to determine predictors for thyroid incidentalomas. Results: In the 1326 patients (mean age 49.4 years and 55.3% female) that were included, the prevalence of thyroid incidentalomas was 20.2%. Patients with thyroid incidentalomas were older (59.6 years vs 46.8 years, P < .001) and more often female than those without incidentalomas (63.4% vs 53.2%, P = .003). On multivariate analysis, only female sex (OR 1.56; 95% CI 1.17-2.07) and older age (OR 1.04; 95% CI 1.03-1.05) were significantly associated with thyroid incidentalomas. Conclusion: In COVID-19 patients, the prevalence of thyroid incidentalomas identified on chest CT was higher (20.2%) than in previous studies in the general population (<1% to 16.8%). Female sex and older age were independent factors associated with thyroid incidentalomas.

Autoimmune hepatitis after vaccination against COVID-19.

In this case report, we present a 61-year-old patient admitted to the hospital because of tiredness and jaundice less than three weeks after vaccination against SARS-CoV-2 with the first dose of the Comirnaty vaccine (Pfizer/ BioNTech). Based on the patient s medical history, laboratory data, imaging methods and liver biopsy, we diagnosed autoimmune hepatitis. The patient developed acute liver failure, and his liver function did not improve after corticosteroid administration. Therefore, the patient was enrolled in the waiting list and underwent a successful orthotopic liver transplantation.Copyright © 2023 Galen s.r.o.. All rights reserved.

Life-threatening Coagulation Disorders

The article discusses the task of radiologic technologists to know clotting disorders and image them best. Topics covered include the various symptoms and blood clots of patients with thrombotic disorders, and medical imaging's beneficial indication of the severity and blood clots' location in the patient's circulatory system, and support for accurate diagnosis and appropriate treatment. Also noted is the boost for positive patient outcomes when the health care team works together.

Comparison of Chest CT Findings between COVID-19 Pneumonia and Influenza Pneumonia

Introduction: With increasing global concerns about the prevalence of COVID-19, chest imaging findings are essential for effective diseases diagnosis and treatment. There is a need to distinguish between imaging features of COVID-19 pneumonia and other viral pneumonia like Influenza.For this purpose, a study was performed on a comparison of chest CT findings between COVID-19 pneumonia and Influenza pneumonia. Method(s): Fifty patients with respiratory symptoms and positive real-time PCR (RT-PCR) of nasopharyngeal swab for Influenza and fifty patients with respiratory symptoms and positive real-time PCR (RT-PCR) of nasopharyngeal swabfor COVID-19 from March to May 2020 were enrolled in the study. In the patient's checklist, information such as demographic characteristics (age, sex), laboratory findings including (CRP, ESR, WBC), and clinical signs (fever, cough, fatigue, dyspnea) were also recorded. Result(s): Gastrointestinal symptoms, anorexia, high CRP, ground-glass opacityare more common in patients with COVID-19 pneumonia than in patients with influenza pneumonia and this difference was statistically significant (P <0.05). But, fever is more common in influenza patients than in Covid-19 patients and this difference is statistically significant (P=0.029). The location of CT scan findings in COVID-19 patients was dominant in peripheral (54 %), while the location of CT scan findings in patients with Influenza was dominant in central (32 %), which is statistically significant (P <0.05). Conclusion(s): According to the results of the study, lung CTscan findings along with some clinical and laboratory findings can help differentiate COVID-19 pneumonia from influenza pneumonia, which is very important in faster diagnosis and timely treatment of both diseases.Copyright © 2023 The Russian Archives of Internal Medicine. All rights reserved.

Vascular Spectrum of Imaging Findings in COVID-19: Ischemic, Hemorrhagic, and Thromboembolic Complications

Abstract Ischemic strokes secondary to occlusion of large vessels have been described in patients with COVID-19. Also, venous thrombosis and pulmonary thromboembolism have been related to the disease. Vascular occlusion may be associated with a prothrombotic state due to COVID-19-related coagulopathy and endotheliopathy. Intracranial hemorrhagic lesions can additionally be seen in these patients. The causative mechanism of hemorrhage could be associated with anticoagulant therapy or factors such as coagulopathy and endotheliopathy. We report on cases of ischemic, thrombotic, and hemorrhagic complications in six patients diagnosed with SARS-CoV-2 infection. Chest computed tomography (CT) showed typical SARS-CoV-2 pneumonia findings in all the cases, which were all confirmed by either serology or reverse transcription polymerase chain reaction (RT-PCR) tests.

Challenges and Burdens in the Coronary Artery Disease Care Pathway for Patients Undergoing Percutaneous Coronary Intervention: A Contemporary Narrative Review.

Clinical and economic burdens exist within the coronary artery disease (CAD) care pathway despite advances in diagnosis and treatment and the increasing utilization of percutaneous coronary intervention (PCI). However, research presenting a comprehensive assessment of the challenges across this pathway is scarce. This contemporary review identifies relevant studies related to inefficiencies in the diagnosis, treatment, and management of CAD, including clinician, patient, and economic burdens. Studies demonstrating the benefits of integration and automation within the catheterization laboratory and across the CAD care pathway were also included. Most studies were published in the last 5-10 years and focused on North America and Europe. The review demonstrated multiple potentially avoidable inefficiencies, with a focus on access, appropriate use, conduct, and follow-up related to PCI. Inefficiencies included misdiagnosis, delays in emergency care, suboptimal testing, longer procedure times, risk of recurrent cardiac events, incomplete treatment, and challenges accessing and adhering to post-acute care. Across the CAD pathway, this review revealed that high clinician burnout, complex technologies, radiation, and contrast media exposure, amongst others, negatively impact workflow and patient care. Potential solutions include greater integration and interoperability between technologies and systems, improved standardization, and increased automation to reduce burdens in CAD and improve patient outcomes.

Prognostic utility of the chest computed tomography severity score for the requirement of mechanical ventilation and mortality in hospitalized patients with COVID-19.

Purpose: To correlate the chest computed tomography severity score (CT-SS) with the need for mechanical ventilation and mortality in hospitalized patients with COVID-19. Materials and methods: The chest CT images of 224 inpatients with COVID-19, confirmed by reverse transcriptase-polymerase chain reaction (RT-PCR), were retrospectively reviewed from April 1 to 25, 2020, in a tertiary health care center. We calculated the CT-SS (dividing each lung into 20 segments and assigning scores of 0, 1, and 2 due to opacification involving 0%, <50%, and ≥50% of each region for a global range of 0-40 points, including both lungs), and collected clinical data. The receiver operating characteristic curve and Youden Index analysis was performed to calculate the CT-SS threshold and accuracy for classification for risk of mortality or MV requirement. Results: 136 men and 88 women were recruited, with an age range of 23-91 years and a mean of 50.17 years; 79 met the MV criteria, and 53 were nonsurvivors. The optimal threshold was >27.5 points for mortality (area under ROC curve >0.96), with a sensitivity of 93% and specificity of 87%, and >25.5 points for the need for MV (area under ROC curve >0.94), with a sensitivity of 90% and specificity of 89%. The Kaplan-Meier curves show a significant difference in mortality by the CT-SS threshold (Log Rank p < 0.001). Conclusions: In our cohort of hospitalized patients with COVID-19, the CT-SS accurately discriminates the need for MV and mortality risk. In conjunction with clinical status and laboratory data, the CT-SS may be a useful imaging tool that could be included in establishing a prognosis for this population.

Abnormal echocardiographic findings after COVID-19 infection: a multicenter registry.

The Coronavirus Disease 2019 (COVID-19) pandemic has transformed health systems worldwide. There is conflicting data regarding the degree of cardiovascular involvement following infection. A registry was designed to evaluate the prevalence of echocardiographic abnormalities in adults recovered from COVID-19. We prospectively evaluated 595 participants (mean age 45.5 ± 14.9 years; 50.8% female) from 10 institutions in Argentina and Brazil. Median time between infection and evaluation was two months, and 82.5% of participants were not hospitalized for their infection. Echocardiographic studies were conducted with General Electric equipment; 2DE imaging and global longitudinal strain (GLS) of both ventricles were performed. A total of 61.7% of the participants denied relevant cardiovascular history and 41.8% had prolonged symptoms after resolution of COVID-19 infection. Mean left ventricular ejection fraction (LVEF) was 61.0 ± 5.5% overall. In patients without prior comorbidities, 8.2% had some echocardiographic abnormality: 5.7% had reduced GLS, 3.0% had a LVEF below normal range, and 1.1% had wall motion abnormalities. The right ventricle (RV) was dilated in 1.6% of participants, 3.1% had a reduced GLS, and 0.27% had reduced RV function. Mild pericardial effusion was observed in 0.82% of participants. Male patients were more likely to have new echocardiographic abnormalities (OR 2.82, p = 0.002). Time elapsed since infection resolution (p = 0.245), presence of symptoms (p = 0.927), or history of hospitalization during infection (p = 0.671) did not have any correlation with echocardiographic abnormalities. Cardiovascular abnormalities after COVID-19 infection are rare and usually mild, especially following mild infection, being a low GLS of left and right ventricle, the most common ones in our registry. Post COVID cardiac abnormalities may be more frequent among males.

Effect of the COVID-19 pandemic on young stroke

Background & Objective: Covid-19 infection has diverse effect on human health. We aimed to evaluate the effect of COVID-19 pandemic on the young stroke cases in an emergency services in a tertiary hospital in Istanbul, Turkey. Method: A total of 86 patients younger than 50 years confirmed to have stroke seen between January 1, 2019 and December 31, 2020 were included in the study. The year 2019 was defined as the pre-pandemic period and the year 2020 as the pandemic period. The patients' stroke type, localization, mortality, laboratory and imaging data were evaluated. Results: Eighty-six patients were included in the study. The mean age was 38.69±5.39 years, 49 (57%) were female. Of the patients, 78 (90.7%) were ischemic and 8 (9.3%) were hemorrhagic stroke. In the pandemic group, ischemic stroke was observed in 55 (96.5%) and hemorrhagic stroke in 2 (3.5%) (p=0.010). While the mean age of the patients in the survival group was 39.24±5.70 years, it was 36.61±3.38 years in the mortality group (p=0.008). While the mortality was 18 (20.9%) overall, it was 16 (18.6%) patients during the pandemic period, and 2 (2.3%) patients in the pre-pandemic period, the difference was statistically significant. (p=0.014). Conclusion: COVID-19 infection appear to increase the risk of ischemic stroke and worsens the mortality among the young. More comprehensive and prospective studies are needed to confirm this observation. © 2023, ASEAN Neurological Association. All rights reserved.

Do Preprocessing and Class Imbalance Matter to the Deep Image Classifiers for COVID-19 Detection? An Explainable Analysis

In a world withstanding the waves of a raging pandemic, respiratory disease detection from chest radiological images using machine-learning approaches has never been more important for a widely accessible and prompt initial diagnosis. A standard machine-learning disease detection workflow that takes an image as input and provides a diagnosis in return usually consists of four key components, namely input preprocessor, data irregularities (like class imbalance, missing and absent features, etc.) handler, classifier, and a decision explainer for better clarity. In this study, we investigate the impact of the three primary components of the disease-detection workflow leaving only the deep image classifier. We specifically aim to validate if the deep classifiers may significantly benefit from additional preprocessing and efficient handling of data irregularities in a disease-diagnosis workflow. To elaborate, we explore the applicability of seven traditional and deep preprocessing techniques along with four class imbalance handling approaches for a deep classifier, such as ResNet-50, in the task of respiratory disease detection from chest radiological images. While deep classifiers are more capable than their traditional counterparts, explaining their decision process is a significant challenge. Therefore, we also employ three gradient visualization algorithms to explain the decision of a deep classifier to understand how well each of them can highlight the key visual features of the different respiratory diseases. © 2020 IEEE.

Computed Tomography Simulator (Ctlab)

University of Oulu and Oulu University of Applied Sciences have established a unique medical imaging teaching and testing laboratory in collaboration with Oulu University Hospital in a European Regional Development Fund -project. Virtually implemented medical imaging devices (CT, MRI, radiography) are unique features of the lab. Many of the virtual tools have been developed by the universities themselves. One of the virtual tools implemented during the project is the CTlab simulator, which can be widely used in computed tomography training for all professionals who use radiation in their work. The CTlab provides fast, comprehensive, and efficient solutions for numerical CT simulations with low hardware requirements. The simulator has been developed to introduce the basic operations and workflow behind the CT imaging modality and to illustrate how the polychromatic x-ray spectrum, various imaging parameters, scan geometry and CT reconstruction algorithm affect the quality of the detected images. Key user groups for the simulator include medical physics, engineering, and radiographer students. CTlab has been created with MATLAB's app designer feature. It offers its user the opportunity to select the virtual imaging target, to adjust CT imaging parameters (image volume, scan angles, detector element size and detector width, noise, algorithm/geometry specific parameters), to select specific scan geometry, to observe projection data from selected imaging target with polychromatic x-ray spectrum, and to select the specific algorithm for image reconstruction (FBP, least squares, Tikhonov regularization). The CTlab has so far been used at a postgraduate course on computed tomography technology with encouraging feedback from the students. At the course, teaching of CT modality were performed by using the simulator, giving students unlimited opportunity to practice the use of virtual imaging device and participate demonstrations remotely during the Covid-19 pandemic. Using CTlab in teaching enhances and deepens the learning experience in the physics behind computed tomography. CTlab can be used remotely (https://www.oulu.fi/fi/projektit/laaketieteellisen-kuvantamisen-opetus-ja-testilaboratorio-0), which makes teaching and training of CT scanner usage successful regardless of time and place. The simulator enables more illustrative and in-depth teaching and offers cost-effectiveness, versatility, and flexibility in education. CTlab can also be used to support teaching in special situations, such as during the Covid-19 pandemic when simulator is utilized remotely to perform teaching-related demonstrations flexibly and safely.Copyright © 2023 Southern Society for Clinical Investigation.

GSAL: Geometric structure adversarial learning for robust medical image segmentation

• A new learning mechanism for medical image segmentation. We introduce a novel Geometric Structure Learning Mechanism (GSLM) that enhances model learning "focus, path, and difficulty". It enables geometric structure attention learning to bridge image features with large differences, thus capturing the contextual dependencies of images. The image features maintain consistency and continuity along the internal and external geometry structure, which improves the integrity and boundary accuracy of the segmentation results. To the best of our knowledge, we are the first attempt to explicitly establish the target's geometric structure, which has been successfully applied to medical image segmentation. • A novel geometric structure adversarial learning for robust medical image segmentation. We present the geometric structure adversarial learning model (GSAL) that consists of a geometric structure generator, skeleton-like and boundary discriminators, and a geometric structure fusion sub-network. The generator yields the geometric structure that preserves interior characteristics consistency and external boundary structure continuity. The dual discriminators are trained simultaneously to enhance and correct the characterization of interior structure and boundary structure, respectively. The fusion sub-network aims to fuse the geometric structure that optimized by adversarial learning to refine the final segmentation results with higher credibility. • State-of-art results on widely-used benchmarks. Our GSAL achieves SOTA performance on a variety of benchmarks, including Kvasir&CVC-612 dataset, COVID-19 dataset, and LIDC-IDRI dataset. It confirms the robustness and generalizability of our framework. In addition, our method has great advantages in terms of the integrity and boundary accuracy of the segmentation target compared to other competitive methods. GSAL can also achieve a considerable trade-off in terms of accuracy, inference speed, and model complexity, which helps deploy in clinical practice systems. Automatic medical image segmentation plays a crucial role in clinical diagnosis and treatment. However, it is still a challenging task due to the complex interior characteristics (e.g. , inconsistent intensity, low contrast, texture heterogeneity) and ambiguous external boundary structures. In this paper, we introduce a novel geometric structure learning mechanism (GSLM) to overcome the limitations of existing segmentation models that lack learning "focus, path, and difficulty." The geometric structure in this mechanism is jointly characterized by the skeleton-like structure extracted by the mask distance transform (MDT) and the boundary structure extracted by the mask distance inverse transform (MDIT). Among them, the skeleton-like and boundary pay attention to the trend of interior characteristics consistency and external structure continuity, respectively. With this idea, we design GSAL, a novel end-to-end geometric structure adversarial learning for robust medical image segmentation. GSAL has four components: a geometric structure generator, which yields the geometric structure to learn the most discriminative features that preserve interior characteristics consistency and external boundary structure continuity, skeleton-like and boundary structure discriminators, which enhance and correct the characterization of internal and external geometry to mutually promote the capture of global contextual dependencies, and a geometric structure fusion sub-network, which fuses the two complementary and refined skeleton-like and boundary structures to generate the high-quality segmentation results. The proposed approach has been successfully applied to three different challenging medical image segmentation tasks, including polyp segmentation, COVID-19 lung infection segmentation, and lung nodule segmentation. Extensive experimental results demonstrate that the proposed GSAL achieves favorably against most state-of-the-art methods under different evaluation metrics. The code is available at: https://github.com/DLWK/GSAL. [ BSTRACT FROM AUTHOR] Copyright of Pattern Recognition is the property of Pergamon Press - An Imprint of Elsevier Science and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full . (Copyright applies to all s.)

Robotic Manipulation System Design and Control for Non-Contact Remote Diagnosis in Otolaryngology: Digital Twin Approach

The COVID-19 pandemic has emphasized the need for non-contact medical robots to alleviate the heavy workload and emotional stress experienced by healthcare professionals while preventing infection. In response, we propose a non-contact robotic diagnostic system for otolaryngology clinics, utilizing a digital twin model for initial design optimization. The system employs a master-slave robot architecture, with the slave robot comprising a flexible endoscope manipulation robot and a parallel robot arm for controlling additional medical instruments. The novel 4 degrees of freedom (DOF) control mechanism enables the single robotic arm to handle the endoscope, facilitating the process compared to the traditional two-handed approach. Phantom experiments were conducted to evaluate the effectiveness of the proposed flexible endoscope manipulation system in terms of diagnosis completion time, NASA task load index (NASA-TLX), and subjective risk score. The results demonstrate the system's usability and its potential to alternate conventional diagnosis. © 2013 IEEE.

Lung Ultrasound versus Chest X-ray for the Diagnosis of COVID-19 Pneumonia

Objectives: The objective of this study is to compare point-of-care lung ultrasound (LUS) with chest x-ray (CXR) to determine which is the more accurate diagnostic imaging modality for COVID-19 pneumonia. Method(s): This was a single-center, prospective, observational IRB-approved study at an urban university hospital with >105,000 patient visits annually. Patients >18 years old, who presented to the emergency department with signs and symptoms of COVID-19, were eligible for enrollment. Each patient received a LUS, performed by an emergency medicine resident or attending physician, using a portable, handheld ultrasound followed by a portable anteroposterior CXR. High risk patients or those with an abnormal imaging finding underwent a non-contrast-enhanced computed tomography (NCCT) as the diagnostic standard. The primary outcome was the sensitivity and specificity of LUS and of CXR at identifying COVID-19 pneumonia against NCCT as the reference standard. Using a power analysis of 80%, our sample size calculation of 98 patients was based on previous data demonstrating a 20% difference in sensitivities between LUS and CXR at diagnosing viral pneumonia. Data are presented as proportions with 95% confidence intervals (CIs). Data analysis included the chi-square and t tests. Background(s): The viral illness, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), commonly known as coronavirus 2019 (COVID-19), has become a global pandemic infecting over 400 million individuals worldwide. Often symptoms are vague, and physical exam findings have proven unreliable as indicators of infection. Therefore, diagnosis typically relies on imaging or nasopharyngeal swabs. Result(s): 143 consecutive patients with signs and symptoms of COVID-19 were approached and enrolled. 27 patients were considered low risk by the attending EP per emergency department guidelines, and 6 patients were admitted for alternate diagnoses without advanced imaging. 110 patients underwent LUS, CXR, and NCCT. 99 LUS and 73 CXRs were interpreted as positive. 81 NCCT were interpreted as positive providing a prevalence of COVID-19 pneumonia of 75% (95% CI 66.0-83.2) in our study population. Sensitivity of LUS was 97.6% (95% CI 91.6-99.7) vs 69.9% (95% CI 58.8-79.5) for CXR. Specificity was 33.3% (95% CI 16.5-54.0) for LUS and 44.4% (95% CI 25.5-64.7) for CXR. LUS positive and negative likelihood ratios were 1.46 (95% CI 1.12-1.92) and 0.0723 (95% CI 0.01-0.31), respectively vs 1.26 (95% CI 0.87-1.81) and 0.67 (95% CI 0.39-1.16) for CXR. PPV and NPV for LUS were 81.8% (95% CI 72.8-88.9) and 81.8% (95% CI 48.2-97.7) compared to 79.5% (95% CI 68.4-88.0) and 32.4% (95% CI 18.0-49.8) for CXR. Conclusion(s): LUS was more sensitive than CXR at identifying COVID-19 pneumonia radiographically. LUS using a portable, handheld ultrasound can be a valuable triage screening modality for patients with suspected COVID-19 pneumonia in diverse clinical settings.Copyright © 2023

Quo vadis Radiomics? Bibliometric analysis of 10-year Radiomics journey.

OBJECTIVES: Radiomics is the high-throughput extraction of mineable and-possibly-reproducible quantitative imaging features from medical imaging. The aim of this work is to perform an unbiased bibliometric analysis on Radiomics 10 years after the first work became available, to highlight its status, pitfalls, and growing interest. METHODS: Scopus database was used to investigate all the available English manuscripts about Radiomics. R Bibliometrix package was used for data analysis: a cumulative analysis of document categories, authors affiliations, country scientific collaborations, institution collaboration networks, keyword analysis, comprehensive of co-occurrence network, thematic map analysis, and 2021 sub-analysis of trend topics was performed. RESULTS: A total of 5623 articles and 16,833 authors from 908 different sources have been identified. The first available document was published in March 2012, while the most recent included was released on the 31st of December 2021. China and USA were the most productive countries. Co-occurrence network analysis identified five words clusters based on top 50 authors' keywords: Radiomics, computed tomography, radiogenomics, deep learning, tomography. Trend topics analysis for 2021 showed an increased interest in artificial intelligence (n = 286), nomogram (n = 166), hepatocellular carcinoma (n = 125), COVID-19 (n = 63), and X-ray computed (n = 60). CONCLUSIONS: Our work demonstrates the importance of bibliometrics in aggregating information that otherwise would not be available in a granular analysis, detecting unknown patterns in Radiomics publications, while highlighting potential developments to ensure knowledge dissemination in the field and its future real-life applications in the clinical practice. CLINICAL RELEVANCE STATEMENT: This work aims to shed light on the state of the art in radiomics, which offers numerous tangible and intangible benefits, and to encourage its integration in the contemporary clinical practice for more precise imaging analysis. KEY POINTS: • ML-based bibliometric analysis is fundamental to detect unknown pattern of data in Radiomics publications. • A raising interest in the field, the most relevant collaborations, keywords co-occurrence network, and trending topics have been investigated. • Some pitfalls still exist, including the scarce standardization and the relative lack of homogeneity across studies.