Cooperative AI training for cardiothoracic segmentation in computed tomography: An iterative multi-center annotation approach.

PURPOSE: Radiological reporting is transitioning to quantitative analysis, requiring large-scale multi-center validation of biomarkers. A major prerequisite and bottleneck for this task is the voxelwise annotation of image data, which is time-consuming for large cohorts. In this study, we propose an iterative training workflow to support and facilitate such segmentation tasks, specifically for high-resolution thoracic CT data. METHODS: Our study included 132 thoracic CT scans from clinical practice, annotated by 13 radiologists. In three iterative training experiments, we aimed to improve and accelerate segmentation of the heart and mediastinum. Each experiment started with manual segmentation of 5-25 CT scans, which served as training data for a nnU-Net. Further iterations incorporated AI pre-segmentation and human correction to improve accuracy, accelerate the annotation process, and reduce human involvement over time. RESULTS: Results showed consistent improvement in AI model quality with each iteration. Resampled datasets improved the Dice similarity coefficients for both the heart (DCS 0.91 [0.88; 0.92]) and the mediastinum (DCS 0.95 [0.94; 0.95]). Our AI models reduced human interaction time by 50 % for heart and 70 % for mediastinum segmentation in the most potent iteration. A model trained on only five datasets achieved satisfactory results (DCS > 0.90). CONCLUSIONS: The iterative training workflow provides an efficient method for training AI-based segmentation models in multi-center studies, improving accuracy over time and simultaneously reducing human intervention. Future work will explore the use of fewer initial datasets and additional pre-processing methods to enhance model quality.

Standardization of a CT Protocol for Imaging Patients with Suspected COVID-19-A RACOON Project.

CT protocols that diagnose COVID-19 vary in regard to the associated radiation exposure and the desired image quality (IQ). This study aims to evaluate CT protocols of hospitals participating in the RACOON (Radiological Cooperative Network) project, consolidating CT protocols to provide recommendations and strategies for future pandemics. In this retrospective study, CT acquisitions of COVID-19 patients scanned between March 2020 and October 2020 (RACOON phase 1) were included, and all non-contrast protocols were evaluated. For this purpose, CT protocol parameters, IQ ratings, radiation exposure (CTDIvol), and central patient diameters were sampled. Eventually, the data from 14 sites and 534 CT acquisitions were analyzed. IQ was rated good for 81% of the evaluated examinations. Motion, beam-hardening artefacts, or image noise were reasons for a suboptimal IQ. The tube potential ranged between 80 and 140 kVp, with the majority between 100 and 120 kVp. CTDIvol was 3.7 ± 3.4 mGy. Most healthcare facilities included did not have a specific non-contrast CT protocol. Furthermore, CT protocols for chest imaging varied in their settings and radiation exposure. In future, it will be necessary to make recommendations regarding the required IQ and protocol parameters for the majority of CT scanners to enable comparable IQ as well as radiation exposure for different sites but identical diagnostic questions.

T2* map at cardiac MRI reveals incidental hepatic and cardiac iron overload.

PURPOSE: The purpose of this study was to assess the diagnostic capabilities of cardiac magnetic resonance (CMR) T2* mapping in detecting incidental hepatic and cardiac iron overload. MATERIALS AND METHODS: Patients with various clinical indications for CMR examination were consecutively included at a single center from January 2019 to April 2023. All patients underwent T2* mapping at 1.5 T in a single mid-ventricular short-axis as part of a comprehensive routine CMR protocol. T2* measurements were performed of the heart (using a region-of-interest in the interventricular septum) and the liver, categorized according to the severity of iron overload. The degree of cardiac iron overload was categorized as mild (15 ms < T2* < 20 ms), moderate (10 ms < T2* < 15 ms) and severe (T2* < 10 ms). The degree of hepatic iron overload was categorized as mild (4 ms < T2* < 8 ms), moderate (2 ms < T2* < 4 ms), severe (T2* < 2 ms). Image quality and inter-reader agreement were assessed using intraclass correlation coefficient (ICC). RESULTS: CMR examinations from 614 patients (374 men, 240 women) with a mean age of 50 ± 18 (standard deviation) years were fully evaluable. A total of 24/614 patients (3.9%) demonstrated incidental hepatic iron overload; of these, 22/614 patients (3.6%) had mild hepatic iron overload, and 2/614 patients (0.3%) had moderate hepatic iron overload. Seven out of 614 patients (1.1%) had incidental cardiac iron overload; of these, 5/614 patients (0.8%) had mild iron overload, 1/614 patients (0.2%) had moderate iron overload, and 1/614 patients (0.2%) had severe iron overload. Good to excellent inter-reader agreement was observed for the assessment of T2* values (ICC, 0.90 for heart [95% confidence interval: 0.88-0.91]; ICC, 0.91 for liver [95% confidence interval: 0.89-0.92]). CONCLUSION: Analysis of standard CMR T2* maps detects incidental cardiac and hepatic iron overload in 1.1% and 3.9% of patients, respectively, which may have implications for further patient management. Therefore, despite an overall low number of incidental abnormal findings, T2* imaging may be included in a standardized comprehensive CMR protocol.

Incidental Bystander or Essential Culprit: A Systematic Review of Bacterial Significance in the Pathogenesis of Breast Implant-Associated Anaplastic Large Cell Lymphoma.

Breast implant-associated anaplastic large cell lymphoma (BIA-ALCL) is a distinct subtype of T-cell non-Hodgkin lymphoma that arises in the context of prolonged exposure to textured breast implants. The intent of this manuscript is to explore whether the bacterial presence in biofilms on these implants is a mere incidental finding or plays a pivotal role in the pathogenesis of BIA-ALCL. Our goal is to delineate the extent of bacterial involvement, offering insights into potential underlying mechanisms, and establishing future research priorities aimed at resolving the remaining uncertainties surrounding this complex association. A comprehensive systematic review of several databases was performed. The search strategy was designed and conducted by an experienced librarian using controlled vocabulary with keywords. The electronic search identified 442 publications. After evaluation, six studies from 2015 to 2021 were included, encompassing 201 female patients aged 23 to 75. The diagnosis span post-implantation ranged from 53 to 135.6 months. Studies consistently found bacteria near breast implants in both BIA-ALCL cases and controls, with varied microbial findings. Both BIA-ALCL cases and controls exhibited the presence of specific bacteria, including Pseudomonas aeruginosa, Klebsiella oxytoca, Staphylococcus aureus, and Ralstonia spp., without any statistically significant differences between groups. The use of antiseptic and antimicrobial agents during implant insertion did not demonstrate any impact on reducing or altering the risk of developing BIA-ALCL. Our systematic review reveals that the current evidence is inadequate to link bacterial etiology as a central factor in the development of BIA-ALCL. The limitations in the existing data prevent a complete dismissal of the role of biofilms in its pathogenesis. The observed gap in knowledge underscores the need for more focused and comprehensive research, which should be structured in a multi-faceted approach. Initially, this involves the utilization of sophisticated genomic and proteomic methods. Following this, it is crucial to delve into the study of immunological reactions specifically induced by biofilms. Finally, this research should incorporate extended observational studies, meticulously tracking the evolution of biofilm development and its correlation with the emergence of BIA-ALCL. In light of the inconclusive nature of current findings, further investigation is not only justified but urgently needed to clarify these unresolved issues.

Perspectives on Participation in Continuous Vocational Education Training-An Interview Study.

In European industrialized countries, a large number of companies in the healthcare, hotel, and catering sectors, as well as in the technology sector, are affected by demographic, political, and technological developments resulting in a greater need of skilled workers with a simultaneous shortage of skilled workers (CEDEFOP, 2015, 2016). Consequently, employers have to address workers who have not been taken into account such as low-skilled workers, workers returning from a career break, people with a migrant background, older people, and jobseekers and train them, in order to guarantee the professionalization of this workforce (Festing and Harsch, 2018). Continuing vocational education and training (CVET) is seen as an indispensable tool; because CVET has advantages for both employers and employees, it helps to increase the productivity of companies (Barrett and O'Connell, 2001), to prevent the widening of socioeconomic disparities (Dieckhoff, 2007), and to open up career opportunities for the workforce (Rubenson and Desjardins, 2009). However, participation rate on CVET seems to differ, depending on institutional factors (such as sector and size of the company) and individual characteristics (such as qualification level, migration background, age and time of absence from work) (e.g., Rubenson and Desjardins, 2009; Wiseman and Parry, 2017). In contrast to previous research, our study aims to provide a holistic view of reasons for and against CVET, combining the different perspectives of employers and (potential) employees. The analysis of reasons and barriers was carried out based on semi-structured interviews. Fifty-seven employers, 73 employees, and 42 jobseekers (potential employees) from the sectors retail, healthcare and social services, hotels and catering, and technology were interviewed. Results point to considerable differences in the reasons and barriers mentioned by the disadvantaged groups. These differences are particularly significant between employees on the one side and employers, as well as jobseekers, on the other side, while the reasons to attend CVET of jobseekers are more similar to those of employers. The results can be used to tailor CVET more closely to the needs of (potential) employees and thus strengthen both the qualification and career opportunities of (potential) employees and the competitiveness and productivity of companies.