Insights into Imaging最新文献

Objectives: Escape rooms provide an interactive learning experience, combining clinical knowledge with problem-solving and teamwork. A radiology-themed escape room has been organized at the European Congress of Radiology in 2019 and 2023-2025, with over 900 people participating in total. The process of developing a radiology-themed escape room is discussed, and the results of a participant survey are presented.

Materials and methods: The development of a radiology-themed escape room was based on five steps. Initially, an overarching concept was chosen, then multiple puzzle ideas were brainstormed. These were linked together to form a story, and then fully developed with relevant images and materials. Finally, the room was tied together, and a fitting atmosphere was created. Participants in 2025 were asked to complete a survey with questions on their training status, the challenges that they found most difficult, and their thoughts on the activity as a learning tool and for improving teamwork.

Results: Three different concepts of radiology-themed escape rooms were developed for the congresses from 2019 to 2025. The overarching concepts were a polytrauma situation, a thrombectomy for fulminant pulmonary embolism, and a tumor board, respectively. Two hundred ninety people participated in 2025, and 149 completed the exit survey; 66.7% of participants were able to complete the room in time. Enjoyment, learning, and team building were all rated highly by participants.

Conclusion: A development process for designing a radiology-themed escape room is presented. A prior implementation shows an enjoyable and educational experience for radiologists and other medical professionals.

Critical relevance: Insights are given on the development of a radiology-themed escape room, providing a unique interactive learning opportunity for residents that incorporates image interpretation with teamwork and cognitive puzzles, resulting in an enjoyable educational experience.

Key points: A step-by-step guide on developing a radiology-themed escape room is presented. Radiological escape rooms provide an enjoyable, educational, and team-building experience. Interactive learning experiences could play a larger role in modern radiology education.

Objective: The purpose of the present study is to determine whether or not lesion characteristics on PET/CT could reduce the number of samples required to achieve a diagnosis in image-guided bone biopsies (IGBB).

Materials and methods: A retrospective review of 38 percutaneous IGBB performed at a single center. Biopsies have been performed from January 1st, 2020, to October 23rd, 2024. Inclusion criteria were patients with a PET/CT and a histopathologic report available. Specimens were collected, numbered, and independently analyzed in separate containers. PET/CT data, including SUV_max, SUV_mean, MTV, TLG, and morphological lesion characteristics, were correlated with biopsy outcomes and subjected to statistical analysis. Patients were classified by the number of samples needed for diagnosis: first (Group 1), second (Group 2), or third/subsequent (Group 3).

Results: Thirty-four/38 (89%) involved spinal and pelvic locations (34/38; 89%). Breast cancer metastases were the most common diagnosis (21/38; 55%). Group 1 included 20 IGBB (52%), group 2 included 9 IGBB (24%), and group 3 included 9 IGBB (24%). No statistically significant difference was found between groups in metabolic characteristics and the number of samples needed for diagnostic purposes (p > 0.05). Subgroup analysis, including factors such as density or lesion size, didn't find any significant differences between groups.

Conclusion: The results suggest that high metabolic activity alone does not justify reducing the number of biopsy samples without compromising diagnostic performance. This supports the recommendation to obtain at least three samples and highlights the importance of selecting the safest biopsy site, regardless of metabolic activity.

Critical relevance statement: This study critically assesses the role of FDG PET/CT metabolic parameters in predicting the diagnostic success of IGBB, providing new insights to improve target selection and biopsy planning in clinical radiology.

Key points: This study assessed whether metabolic activity on FDG PET/CT influences the diagnostic yield of IGBB. High metabolic activity did not allow for reducing the number of samples without affecting diagnostic performance. At least three biopsy samples should be obtained, prioritizing safety over metabolic activity when selecting the biopsy site.

Objectives: In this study, we developed a multi-modal CT-based machine learning model to predict the response of gastric cancer (GC) patients to first-line chemotherapy combined with PD-1 inhibitors and performed external validation and multi-model comparisons.

Materials and methods: We retrospectively analyzed the clinical data of 348 patients with GC who underwent immunotherapy. The patients were categorized into an internal validation cohort (center A, n = 272) and an external validation cohort (center B, n = 76). Pre-treatment clinical and CT radiomics features were extracted to develop three models: a clinical model, a radiomics model and a clinical-radiomics model. The classifiers included logistic regression (LR), linear support vector classification (Linear SVC), support vector machine, and random forest.

Results: A total of 19 radiomics signatures and 5 clinical feature signatures were selected. In the radiomics model, the Linear SVC algorithm achieved an area under the receiver operating characteristic curve (AUC) of 0.88 and 0.76 in internal and external validation sets, respectively. In both the clinical model and the clinical-radiomics model, the LR algorithm demonstrated high and stable predictive performance in the internal (AUC = 0.89 and 0.94) and external validation datasets (AUC = 0.76 and 0.85). Among all models in the external validation dataset, the clinical-radiomics model utilizing LR outperformed all other classifiers.

Conclusions: The clinical-radiomics model, in combination with the LR algorithm, provides a reliable and effective method for predicting the early response of advanced GC patients treated with programmed cell death-1 (PD-1) inhibitors combined with chemotherapy.

Critical relevance statement: CT radiomics and laboratory parameters were used to evaluate early prediction of response to PD-1 inhibitors combined with chemotherapy in patients with advanced gastric cancer. This clinical-radiomics model provides a novel approach to predict immunotherapy efficacy and prognosis.

Key points: Evaluating the efficacy of PD-1 inhibitors combined with chemotherapy in advanced gastric cancer using only clinical data is limited. Only some patients with advanced gastric cancer treated with the PD-1 inhibitors combined with chemotherapy achieved complete regression. This clinical-radiomics model showed good performance for predicting gastric cancer response to chemotherapy combined with PD-1 inhibitors.

Objectives: AI offers considerable potential to improve diagnostic accuracy and efficiency in radiology. However, its successful implementation depends largely on the trust and acceptance of referring physicians. This study examines physicians' attitudes toward AI in radiology, identifying key facilitators and barriers to its clinical integration.

Materials and methods: A total of 169 licensed physicians in Germany, including surgeons, internists, and general practitioners who frequently refer patients to radiology, were surveyed. Participants were recruited via a systematic review of hospital and practice websites. A structured online questionnaire assessed perceptions of AI, focusing on trust-related factors, preferred applications, and adoption barriers. Statistical analysis was conducted using R and Python.

Results: Overall, 60% of respondents evaluated AI positively for enhancing diagnostic accuracy (mean score 3.7 ± 1.2). The most influential trust factor was model transparency (56.3%), followed by legal clarity on liability (25.0%) and strong data protection (11.7%). Transparency was rated significantly higher than other factors (p < 0.001). Preferred AI applications included lesion detection, research data analysis, and workflow management. Barriers to adoption included the "black box" nature of AI, unclear accountability, and data privacy concerns. Subgroup analysis revealed no significant variation in trust factors between specialties (p = 0.21).

Conclusion: Physicians see AI as a promising tool in radiology but emphasize the need for greater transparency, clear legal responsibility, and secure data handling. Addressing these concerns through explainable AI models, legal frameworks, and robust data protection measures is essential for fostering trust and facilitating successful AI integration in clinical practice.

Critical relevance statement: Understanding physicians' concerns about AI transparency, liability, and data privacy is essential. Addressing these barriers is critical to ensuring responsible implementation, building trust, and enabling effective integration of AI into clinical radiology workflows.

Key points: AI acceptance in radiology faces transparency and liability concerns. Lesion detection and data analysis were rated most beneficial by physicians. Clear regulation and explainability are key for clinical AI trust.

Objectives: To prospectively evaluate the diagnostic accuracy of ultrasound-derived fat fraction (UDFF) in quantifying hepatic steatosis, to establish and validate a dual-threshold UDFF classification system, and to investigate its efficacy for risk stratification in body mass index (BMI)-defined subgroups.

Materials and methods: This prospective multicenter study involved 790 suspected metabolic dysfunction-associated steatotic liver disease (MASLD) participants from April 2023 to November 2024 (derivation: n = 553; validation: n = 237). Liver biopsy histopathology (n = 342), MRI proton density fat fraction (MRI-PDFF) (n = 396), or proton magnetic resonance spectroscopy (¹H-MRS) (n = 52) was used as the reference standard. UDFF was compared to noninvasive test Hepatic Steatosis Index (HSI) and Fatty Liver Index (FLI) using area under the curve (AUC). The diagnostic thresholds were optimized to maintain at least 90% sensitivity and specificity in stratifying hepatic steatosis severity. A two-step strategy of UDFF followed by HSI was used to rule in and rule out steatosis at BMI ≥ 23 kg/m² subgroup.

Results: UDFF demonstrated significant correlations with three reference standards (Spearman's ρ = 0.798-0.847). Comparing with HSI and FLI, UDFF showed higher AUC (0.933, 0.948, and 0.914, respectively) for assessing ≥ S1, ≥ S2 and S3. A clinically practical dual-threshold system effectively classified hepatic steatosis severity. A sequential UDFF/HSI strategy achieved a high positive predictive value (PPV = 95.8%) to rule in hepatic steatosis and lowered the proportion of indeterminate cases (from 18.0 to 7.6%) in patients with BMI ≥ 23 kg/m².

Conclusion: UDFF is a highly effective noninvasive tool for quantifying hepatic steatosis. A sequential use of UDFF/HSI could improve hepatic steatosis detection in patients with BMI ≥ 23 kg/m².

Critical relevance statement: The study proposed dual-threshold diagnostic criteria (sensitivity/specificity ≥ 90%) of UDFF for steatosis grading, and established a BMI-stratified risk stratification tool in multi-center cohorts, proving the efficacy of UDFF in noninvasively quantifying liver steatosis.

Key points: Early diagnosis of hepatic steatosis holds critical clinical significance. The study proposed dual-threshold ultrasound-derived fat fraction (UDFF) criteria and BMI-stratified steatosis risk prediction strategy. UDFF provided a non-invasive, accurate diagnostic alternative to liver biopsy and MRI.

Objectives: To evaluate the potential value of ¹⁸F-FDG positron emission tomography (PET) and multiparametric MRI (intravoxel incoherent motion, IVIM, and diffusion kurtosis imaging, DKI) in the prediction of lymphovascular invasion (LVI) in non-small cell lung cancer (NSCLC).

Materials and methods: A total of 73 patients with NSCLC who underwent integrated ¹⁸F-FDG PET/MRI were included. IVIM, DKI, and PET parameters with or without LVI of NSCLC were measured and compared, and the area under the receiver operating characteristic curve (AUC) was used to evaluate the diagnostic efficacy of each parameter. Univariate and multivariate logistic regression models were used to study the optimal combination of PET/MRI parameters for predicting LVI.

Results: PET-derived parameters (SUVmax, MTV, TLG) and IVIM, DKI MRI-derived parameters (ADCstand, D, MK, MD) were significantly different between patients with and without LVI (p < 0.05). Multivariate logistic regression analysis showed that MTV and D were independent predictors of LVI, and the combined prediction model of the two parameters had the highest predictive value for the diagnosis of LVI (AUC = 0.841; sensitivity = 63.83%; specificity = 92.31%).

Conclusion: The present study demonstrates that IVIM, DKI, and PET can be utilized to evaluate LVI status in NSCLC, with the combined diagnostic approach of MTV and D showing the highest diagnostic performance, which may provide a novel reference for clinical management.

Critical relevance statement: The performance of metabolic parameters and diffusion parameters in the identification of lymphovascular invasion (LVI) in non-small cell lung cancer (NSCLC) is similar, but the combination of metabolic tumor volume (MTV) and true diffusion coefficient (D) may improve the diagnostic efficacy.

Key points: A multimodal PET-MRI model evaluates lymphovascular invasion (LVI) in patients with non-small cell lung cancer (NSCLC). Metabolic and diffusion parameters have similar efficacy in predicting LVI in NSCLC. The combined metabolic tumor volume and true diffusion coefficient prediction model is the most valuable.

Objectives: To evaluate gender representation among editors-in-chief and deputy editors of radiology journals indexed in the 2024 Journal Citation Reports (JCR) and to analyze associations with bibliometric indicators and global economic classification.

Materials and methods: A cross-sectional study was performed using publicly available data from radiology-related journals listed in the 2024 JCR (released June 2025). Journals were included if the editorial board composition was accessible online. Gender was identified through institutional profiles and standardized databases. Descriptive statistics summarized gender distribution. Associations between gender, editorial role, bibliometric performance, and World Bank income classification were tested using chi-square, Mann-Whitney U, Spearman's correlation, and nominal logistic regression.

Results: Of 204 eligible journals, 135 met the inclusion criteria, comprising 387 editorial members. Women represented 20.2% of all editors, 21.4% of deputy editors, and 18.4% of editors-in-chief. Female representation was highest in Q1 journals (26.0%) and lowest in Q2 (15.1%). A significant association was observed between Eigenfactor Score and female representation (p = 0.0494), whereas no association was found with journal impact factor or income classification. Geographic disparities were evident, with some countries achieving parity while others had no female representation.

Conclusions: Gender inequities remain pronounced in radiology editorial leadership, particularly at the editor-in-chief level. Higher Eigenfactor Scores may modestly correlate with improved inclusion. Transparent policies and targeted interventions are required to address structural inequities and advance diversity in academic publishing.

Critical relevance statement: Gender disparities exist in radiology editorial leadership, and the Eigenfactor Score was found to be associated with female representation. By providing a comprehensive overview, the findings underscore the structural barriers that limit diversity and the importance of transparent, equity-focused editorial policies.

Key points: Gender disparities persist in radiology editorial boards, with women underrepresented at both deputy editor and editor-in-chief levels. Eigenfactor Score, but not impact factor or national income classification, was significantly associated with increased female representation. Gender disparities persist across editorial leadership roles in radiology, underscoring the need for transparent policies and structural reforms to promote greater equity.

Objectives: To compare understandability, readability, and communication characteristics of radiology report explanations generated by three freely accessible large language models-ChatGPT, Gemini, and Copilot-based on a standardized prompt, as assessed by expert reviewers.

Materials and methods: In this retrospective single-center study, 100 anonymized radiology reports were randomly selected from five subspecialties. Each report was submitted to ChatGPT (GPT-3.5), Gemini, and Copilot between May 23 and May 30, 2025, using the prompt, "Can you explain my radiology report?". Responses were evaluated for medical correctness on a 3-point scale (0-2), understandability using the patient education materials assessment tool for understandability (PEMAT-U), and readability using Flesch Reading Ease (FRE), Automated Readability Index (ARI), and Gunning Fog Index (GFI). Communicative features-including uncertainty language, patient guidance, and clinical suggestions-were also assessed. Anxiety-inducing potential was rated on a 3-point Likert scale.

Results: All models demonstrated high medical correctness (mean: 1.97 ± 0.17/2). ChatGPT produced the most readable (FRE: 60.33 ± 3.65; ARI: 9.66 ± 1.01; GFI: 9.1 ± 1.04) and understandable (PEMAT-U: 89.58 ± 3.90%) responses (p < 0.01). Copilot included the most uncertainty language (1.62 ± 0.62) and clinical suggestions (1.69 ± 0.60), while Gemini provided the strongest patient guidance (1.62 ± 0.58) (all p < 0.01). Only Copilot showed subspecialty-related variation in readability (GFI; p = 0.048). Anxiety potential was low across all models (mean: 0.07 ± 0.33).

Conclusion: ChatGPT offered superior readability and understandability. Copilot delivered more clinical detail with cautious language, while Gemini emphasized patient-centered guidance. These differences support context-specific use of language models in radiology communication.

Critical relevance statement: This study shows that freely accessible large language models produce radiology report explanations with varying levels of readability, understandability, and communication quality. Expert-based findings may help inform future strategies to optimize patient-facing applications of AI in radiological communication.

Key points: This study compared how freely available AI chatbots respond to patient queries about radiology reports. Significant differences were found in understandability, readability, patient guidance, and use of uncertainty or clinical suggestions. Findings support context-specific use of AI tools to improve radiology communication and patient understanding.

Objectives: Multi-material decomposition (MMD), a key application of dual-energy computed tomography (DECT), has shown potential in musculoskeletal research. This study aimed to compare the diagnostic performance of DECT-based MMD with standard CT and MRI in detecting rotator cuff disease.

Materials and methods: This retrospective study evaluated patients diagnosed with rotator cuff disease who underwent third-generation dual-source DECT and 3.0-T MRI within a 2-week interval between December 2023 and November 2024. Shoulder arthroscopy served as the reference standard. Six readers independently assessed rotator cuff tears and determined the degree of supraspinatus tendon diseases using standard CT, DECT-based MMD and MRI. Area under the curve (AUC), sensitivity, specificity, positive/negative predictive values and accuracy were calculated for the diagnosis of rotator cuff disease. Friedman test was used to analyze the radiologists' diagnostic confidence across the three image types.

Results: In total of 103 patients (mean age: 50.0 ± 15.6 years) underwent shoulder arthroscopy. MMD demonstrated a higher average AUC for diagnosing rotator cuff tears (88% vs. 65%, p < 0.001) and supraspinatus tendon disease (86% vs. 63%, p < 0.001) compared to standard CT. Its diagnostic performance for supraspinatus tendon disease (91% vs. 90%, p = 0.35) and full-thickness tears (95% vs. 93%, p = 0.11) was comparable to that of MRI.

Conclusion: DECT-based MMD demonstrated superior diagnostic performance and reliability for detecting rotator cuff diseases compared to standard CT, with accuracy comparable to that of MRI in detecting supraspinatus tendon tears. DECT-based MMD offers a promising approach for the opportunistic detection of rotator cuff diseases.

Critical relevance statement: Dual energy CT-based multi-material decomposition demonstrated accuracy comparable to that of MRI in detecting supraspinatus tendon tears, and may provide an alternative for patients with contraindications to MRI, facilitating early detection of injuries and accurate diagnosis of rotator cuff diseases.

Key points: Dual energy (DE) CT multi-material decomposition (MMD) improves diagnostic performance for rotator cuff tears and supraspinatus tendon injuries. Radiologists with varying experience levels benefited from MMD, with experienced readers achieving MRI-level diagnostic performance. DECT MMD offers a promising alternative for patients with contraindications for MRI.

Background: Research activities often compete with clinical work and personal life for the time of physician-scientists. To overcome barriers to research, examining the factors affecting research productivity is important.

Objectives: To identify potential personal, physician-dependent, and external physician-independent factors affecting researcher productivity in a cohort of Nordic radiologists.

Methods: A prospective survey was open to responders from 10 May 2023 to 23 June 2023. The survey was distributed to radiologists and radiology residents in the Nordic countries (Denmark, Finland, Norway, and Sweden) through multiple channels. We collected demographic information, details about work and academic careers, and opinions and attitudes on work, research, and personal life using a Likert-scale questionnaire.

Results: A total of 192 participants responded (mean age 46.4 (SD: 11.03), 88 (45.8%) males, 103 (53.6%) females). Of the 134 (69.8%) respondents who reported having made any past academic contribution, 88 (46.4%) indicated active research participation. Active researchers expressed more agreement that they have the skills (p < 0.001) and resources (p < 0.001) for research and are able to maintain expertise (p = 0.003). Responders most frequently reported that having time for research (n = 94/155, 60.6%), motivation (n = 56/155, 36.1%), more funding (n = 36/155, 23.2%), and a higher salary (n = 36/155, 23.2%) would increase research involvement.

Conclusions: We identified several differences between radiologists who are active in research and those who are not. The participants identified time, financial means, and motivation as key factors that could increase research involvement.

Critical relevance statement: Academic radiologists with active research careers report having the necessary skills and resources for research, teaching, and learning more frequently than radiologists less active in research.

Key points: Active researchers are more in agreement with having the skills and resources for research. Active researchers expressed more agreement with interest in teaching, keeping expertise up to date, getting enough sleep, and being less distracted by social media. Researchers with a recent history of funding reported more publications. There may be research potential for physicians engaged in self-financed research.