Deep Learning for Pneumothorax Detection on Chest Radiograph: A Diagnostic Test Accuracy Systematic Review and Meta Analysis.

BACKGROUND: Pneumothorax is a common acute presentation in healthcare settings. A chest radiograph (CXR) is often necessary to make the diagnosis, and minimizing the time between presentation and diagnosis is critical to deliver optimal treatment. Deep learning (DL) algorithms have been developed to rapidly identify pathologic findings on various imaging modalities. PURPOSE: The purpose of this systematic review and meta-analysis was to evaluate the overall performance of studies utilizing DL algorithms to detect pneumothorax on CXR. METHODS: A study protocol was created and registered a priori (PROSPERO CRD42023391375). The search strategy included studies published up until January 10, 2023. Inclusion criteria were studies that used adult patients, utilized computer-aided detection of pneumothorax on CXR, dataset was evaluated by a qualified physician, and sufficient data was present to create a 2 × 2 contingency table. Risk of bias was assessed using the QUADAS-2 tool. Bivariate random effects meta-analyses and meta-regression modeling were performed. RESULTS: Twenty-three studies were selected, including 34 011 patients and 34 075 CXRs. The pooled sensitivity and specificity were 87% (95% confidence interval, 81%, 92%) and 95% (95% confidence interval, 92%, 97%), respectively. The study design, use of an institutional/public data set and risk of bias had no significant effect on the sensitivity and specificity of pneumothorax detection. CONCLUSIONS: The relatively high sensitivity and specificity of pneumothorax detection by deep-learning showcases the vast potential for implementation in clinical settings to both augment the workflow of radiologists and assist in more rapid diagnoses and subsequent patient treatment.

Interpregnancy weight change and risks of stillbirth and infant mortality: a protocol of a systematic review and meta-analysis.

INTRODUCTION: Interpregnancy weight change may impact two important adverse perinatal outcomes: stillbirth and infant mortality. This systematic review aims to synthesise the existing evidence on the association between interpregnancy weight change and stillbirth and infant mortality. METHODS AND ANALYSIS: This systematic review and meta-analysis will be conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses Protocols guidelines and has been registered in the International Prospective Register of Systematic Reviews (PROSPERO). A comprehensive literature search of four online databases (Embase, Cochrane Libraries, Web of Science and Medline) will be conducted from inception to October 2023. Observational (longitudinal, cohort, case-control) and randomised controlled trials will be included. Interpregnancy weight/body mass index change between two consecutive pregnancies will be the exposure. The primary outcomes will be the incidence of stillbirth and infant mortality in subsequent pregnancy. The Cochrane Risk of Bias tool will be used to assess the risk of bias in the randomised controlled studies and the Risk of Bias in Non-Randomised Studies of Interventions tool will be used for observational studies. If there are sufficient data, a meta-analysis will be conducted to estimate the pooled effect size. Otherwise, qualitative descriptions of individual studies will be summarised. The heterogeneity will be statistically assessed using a χ2 test and I2 statistic. ETHICS AND DISSEMINATION: Ethics approval is not required for this study as all results will be based on published papers. No primary data collection will be needed. Study findings will be presented at scientific conferences or published in a peer-reviewed scientific journal. TRIAL REGISTRATION NUMBER: A registration for this review has been submitted to PROSPERO under CRD42020222977.

Cross-Sectional Evaluation of Open Science Practices at Imaging Journals: A Meta-Research Study.

Objective: To evaluate open science policies of imaging journals, and compliance to these policies in published articles. Methods: From imaging journals listed we extracted open science policy details: protocol registration, reporting guidelines, funding, ethics and conflicts of interest (COI), data sharing, and open access publishing. The 10 most recently published studies from each journal were assessed to determine adherence to these policies. We calculated the proportion of open science policies into an Open Science Score (OSS) for all journals and articles. We evaluated relationships between OSS and journal/article level variables. Results: 82 journals/820 articles were included. The OSS of journals and articles was 58.3% and 31.8%, respectively. Of the journals, 65.9% had registration and 78.1% had reporting guideline policies. 79.3% of journals were members of COPE, 81.7% had plagiarism policies, 100% required disclosure of funding, and 97.6% required disclosure of COI and ethics approval. 81.7% had data sharing policies and 15.9% were fully open access. 7.8% of articles had a registered protocol, 8.4% followed a reporting guideline, 77.4% disclosed funding, 88.7% disclosed COI, and 85.6% reported ethics approval. 12.3% of articles shared their data. 51% of articles were available through open access or as a preprint. OSS was higher for journal with DOAJ membership (80% vs 54.2%; P < .0001). Impact factor was not correlated with journal OSS. Knowledge synthesis articles has a higher OSS scores (44.5%) than prospective/retrospective studies (32.6%, 30.0%, P < .0001). Conclusion: Imaging journals endorsed just over half of open science practices considered; however, the application of these practices at the article level was lower.