In this review, we focused on the applicability of artificial intelligence (AI) for opportunistic abdominal aortic aneurysm (AAA) detection in computed tomography (CT). We used the academic search system PubMed as the primary source for the literature search and Google Scholar as a supplementary source of evidence. We searched through 2 February 2022. All studies on automated AAA detection or segmentation in noncontrast abdominal CT were included. For bias assessment, we developed and used an adapted version of the QUADAS-2 checklist. We included eight studies with 355 cases, of which 273 (77%) contained AAA. The highest risk of bias and level of applicability concerns were observed for the “patient selection” domain, due to the 100% pathology rate in the majority (75%) of the studies. The mean sensitivity value was 95% (95% CI 100–87%), the mean specificity value was 96.6% (95% CI 100–75.7%), and the mean accuracy value was 95.2% (95% CI 100–54.5%). Half of the included studies performed diagnostic accuracy estimation, with only one study having data on all diagnostic accuracy metrics. Therefore, we conducted a narrative synthesis. Our findings indicate high study heterogeneity, requiring further research with balanced noncontrast CT datasets and adherence to reporting standards in order to validate the high sensitivity value obtained.

Rationale and Objectives: Post-COVID condition (PCC) is associated with long-term neuropsychiatric symptoms. Magnetic resonance imaging (MRI) in PCC examines the brain metabolism, connectivity, and morphometry. Such techniques are not easily available in routine practice. We conducted a scoping review to determine what is known about the routine MRI findings in PCC patients. Materials and Methods: The PubMed database was searched up to 11 April 2023. We included cohort, cross-sectional, and before–after studies in English. Articles with only advanced MRI sequences (DTI, fMRI, VBM, PWI, ASL), preprints, and case reports were excluded. The National Heart, Lung, and Blood Institute and PRISMA Extension tools were used for quality assurance. Results: A total of 7 citations out of 167 were included. The total sample size was 451 patients (average age 51 ± 8 years; 67% female). Five studies followed a single recovering cohort, while two studies compared findings between two severity groups. The MRI findings were perivascular spaces (47%), microbleeds (27%) and white matter lesions (10%). All the studies agreed that PCC manifestations are not associated with specific MRI findings. Conclusion: The results of the included studies are heterogeneous due to the low agreement on the types of MRI abnormalities in PCC. Our findings indicate that the routine brain MRI protocol has little value for long COVID diagnostics.

Introduction . Artificial intelligence (AI) is an effective tool for automating routine tasks in radiology. The diagnostic accuracy of AI in detecting various pathologies on medical images has generated considerable interest in the scientific community: the number of studies and meta-analyses has been constantly growing. The abundance of published evidence and the diversity of outcomes necessitate the need to systematize the available publications. The aim of this paper is to conduct an umbrella systematic review of contemporary meta-analyses on the use of AI in radiology. Materials and methods. PubMed was searched for studies published in the English language. Thirty-eight systematic reviews with meta-analyses published between 2021 and 2023 were selected for full-text analysis. The extracted data included the goal, study design, imaging modality, sample size, quality assessment of the included studies, AI diagnostic accuracy estimates, reference method parameters, and clinical efficacy metrics of AI implementation. The methodological quality of included systematic reviews was assessed using the AMSTAR-2 tool. Results. Nearly half (47%) of the included meta-analyses focused on the diagnosis, staging and segmentation of malignancies. Four meta-analyses were related to detection of maxillofacial structures in dentistry, while another four meta-analyses addressed the diagnosis of brain lesions. The diagnosis of COVID-19 and the diagnosis of bone fractures were each covered in three meta-analyses. One meta-analysis was reviewed for each of the following fields: colorectal polyps, pneumothorax, pulmonary embolism, osteoporosis, aneurysms, multiple sclerosis, acute cerebrovascular accident, intracranial hemorrhage, burns, and the risk of intrauterine growth restriction. Thirty-five (92%) meta-analyses assessed the risk of bias. Twenty-eight (80%) meta-analyses utilized QUADAS-2 to assess the risk of bias. 14 out of 28 papers reported low risk of bias (50%); 4 (14%) – moderate; 10 (36%) – high. The major risks were associated with samples that were unbalanced in terms of size and composition, a lack of details about the methods, a low number of prospective studies, and a lack of external validation of the outcomes. The overall results indicate that the diagnostic accuracy of AI is comparable to or even greater than that of radiologists. The mean sensitivity, specificity and area under the ROC curve for AI and radiologists were 85.2%, 89.5%, 93.5% and 84.4%, 90.0%, 92.8%, respectively. However, many studies that compared the diagnostic accuracy of AI and radiologists lack the data on the number and experience of the latter. Only one paper presented results of implementing AI into routine clinical diagnosis. Discussion. AI is capable of reducing the turnaround time for non-urgent examinations. When used to verify the primary interpretation, AI was effective in detecting false-negative results from radiologists. However, the efficacy of detecting false-positive results was inadequate. Our assessment of the quality of systematic reviews with AMSTAR-2 show that the methods of searching, selecting and analyzing literature must be improved and brought to a common standard. The development of a specialized tool for assessing the quality of systematic reviews in the AI implementation is also necessary. Due to high diagnostic accuracy, AI is currently considered a promising tool for optimizing the turnaround time. However, more evidence is needed to study the AI outcomes in routine clinical practice. Furthermore, it is necessary to standardize and improve the quality of research methodology.

Background/Objectives: Large language models (LLMs) are increasingly considered for use in radiology, including the summarization of patient medical records to support radiologists in processing large volumes of data under time constraints. This task requires not only accuracy and completeness but also clinical applicability. Automatic metrics and general-purpose questionnaires fail to capture these dimensions, and no standardized tool currently exists for the expert evaluation of LLM-generated summaries in radiology. Here, we aimed to develop and validate such a tool. Methods: Items for the questionnaire were formulated and refined through focus group testing with radiologists. Validation was performed on 132 LLM-generated summaries of 44 patient records, each independently assessed by radiologists. Criterion validity was evaluated through known-group differentiation and construct validity through confirmatory factor analysis. Results: The resulting seven-item instrument, ELEGANCE (Expert-Led Evaluation of Generative AI Competence and Excellence), demonstrated excellent internal consistency (Cronbach’s α = 0.95). It encompasses seven dimensions: relevance, completeness, applicability, falsification, satisfaction, structure, and correctness of language and terminology. Confirmatory factor analysis supported a two-factor structure (content and form), with strong fit indices (RMSEA = 0.079, CFI = 0.989, TLI = 0.982, SRMR = 0.029). Criterion validity was confirmed by significant between-group differences (p < 0.001). Conclusions: ELEGANCE is the first validated tool for expert evaluation of LLM-generated medical record summaries for radiologists, providing a standardized framework to ensure quality and clinical utility.

Competencies in statistical data processing are becoming increasingly important for modern scientists. The apparent advantages of open-source software for statistical analysis are its accessibility and adaptability. The programming language and the corresponding software R, available as a minimalistic console interface or a complete development environment RStudio/Posit, have the widest possibilities among free solutions. We present a practical guide for comparing two groups using the software R. This study compares the effective doses of standard computed tomography with low-dose computed tomography for COVID-19 patients. The practical guide summarizes theoretical approaches to medical data processing and recommendations for correctly formulating research tasks and selecting optimal statistical analysis methods. The main goal of the practical guide is to introduce the reader to the Posit interface and the basic functionality of the R language by using a practical example of treating a real medical problem. The presented material can be useful as an introduction to statistical analysis using the programming language R.