Objective: To evaluate the accuracy and reasoning capabilities of large multimodal language models compared with those of neuroradiology subspecialty-trained radiologists in neuroradiology case interpretation. Materials and Methods: This experimental study used custom-made 401 radiologic quizzes derived from articles published in RadioGraphics covering neuroradiology and head and neck topics (October 2020 to February 2024). We prompted the GPT-4 Turbo with Vision (GPT-4V), GPT-4 Omni, Gemini Flash, and Claude models to provide the top three differential diagnoses with a rationale and describe examination characteristics such as imaging modality, sequence, use of contrast, image plane, and body part. The temperature was adjusted to 0 and 1 (T1). Two neuroradiologists answered the same questions.The accuracies of the large language models (LLMs) and the neuroradiologists were compared using generalized estimating equations. Three neuroradiologists assessed the rationale provided by the LLMs for their differential diagnoses using four-point scales, separately for specific lesion locations and imaging findings, and evaluated the presence of hallucinations and the overall acceptability of the responses. Results: Top-3 accuracy (i.e., correct answers present among top-3 differential diagnoses) of LLMs ranged from 29.9% (120 of 401) to 49.4% (198 of 401, obtained with GPT-4V in the T1 setting), while radiologists achieved 80.3% (322 of 401) and 68.3% (274 of 401), respectively (P < 0.001). Regarding the rationale for differential diagnoses, GPT-4V (T1) accurately identified both the specific lesion location and imaging findings in 30.7% (123 of 401) and 12.9% (16 of 124) of cases without textual clinical history. Hallucinations occurred in 4.5% (18 of 401), and only 29.4% (118 of 401) of the LLM-generated analyses were deemed acceptable. GPT-4V (T1) demonstrated high accuracy in identifying the imaging modality (97.4% [800 of 821]) and scanned body parts (92.2% [756 of 820]). Conclusion LLMs remarkably underperformed compared with neuroradiologists and showed unsatisfactory reasoning for their differential diagnoses, with performance declining further in cases without textual input of clinical history. These findings highlight the limitations of current multimodal LLMs in neuroradiological interpretation and their reliance on text input.

Purpose: Radiotherapy (RT) remains a cornerstone of curative treatment for localized and locally advanced prostate cancer. However, dose escalation to improve tumor control is often constrained by the proximity of the rectum, which increases the risk of gastrointestinal (GI) and genitourinary toxicities. Biodegradable rectal spacers inserted between the prostate and rectum have emerged as an effective approach to reduce rectal radiation exposure. This guideline provides evidence-based recommendations on indications, contraindications, procedural standards, and clinical management for biodegradable rectal spacer insertion during prostate cancer RT. Materials and Methods: This guideline was developed by a multidisciplinary expert panel through a systematic review of the literature, analysis of international guidelines (National Comprehensive Cancer Network, European Association of Urology, American Society for Radiation Oncology), and expert consensus among radiation oncologists, radiologists, and urologists with clinical experience in spacer insertion. The strength of each recommendation and the level of evidence were classified according to the modified GRADE (Grading of Recommendations, Assessment, Development, and Evaluation) system. Results: Spacer insertion is conditionally recommended (Grade C, Level I) for patients receiving definitive external-beam RT without rectal invasion. It reduces the high-dose rectal irradiation volume (V70–75) by >50%, decreases acute GI toxicity, and helps maintain bowel-related quality of life. However, the benefit for late severe toxicity (grade 2 or higher) remains debated in recent meta-analyses. Contraindications include rectal invasion, anatomical inaccessibility, infection, and material hypersensitivity. Procedures should be performed under local anesthesia in a sterile environment by trained physicians. Short-course antibiotics and simulator-based training, including completion of multiple supervised cases, are advised. Conclusion Biodegradable rectal spacer insertion is clinically validated and effective in reducing acute rectal toxicity. Although pivotal trials demonstrated a favorable procedural safety profile, real-world postmarket data include reports of rare but severe procedural complications. This guideline provides standardized recommendations tailored to Korean clinical practice while remaining consistent with international standards, emphasizing the importance of operator training and careful patient selection.

Background: The heterogeneity of chronic kidney disease (CKD) and fragmented analysis methods hinder the precise identification of novel biomarkers. We addressed this challenge using two independent cohorts to integrate genomics and metabolomics, aiming to identify cause-specific biomarkers for CKD in the Korean population. Methods: A longitudinal genome-wide association study using the Cox proportional hazards model was conducted using the Ansan and Ansung cohort. To validate these genomic biomarkers and integrate them with plasma metabolomics biomarkers, we utilized a hospital-based biopsy cohort to identify cause-specific CKD biomarkers. Within the biopsy cohort, we analyzed four disease subsets, including type 2 diabetic kidney disease (DKD), hypertensive nephropathy (HN), immunoglobulin A nephropathy (IgAN), and membranous nephropathy (MN), and compared them with healthy individuals. Significant single nucleotide polymorphisms(SNPs) and metabolites for each CKD subset were identified through logistic regression and correlation-based network analyses. Subsequently, we analyzed the risk of disease progression associated with the identified pairs. Results: A total of 448 variants associated with CKD occurrence were identified, with significant differences in several genetic variants and metabolites observed among patients with DKD, HN, IgAN, and MN compared to healthy individuals. Among 36 SNP-metabolite pairs, those involing FOXB1 and ZFP42 were associated with DKD, whereas pairs involving MMRN1 and SYNJ2 were linked to MN. Notably, the rs1025170 variant in FOXB1 and tyrosine pair was correlated with DKD progression. Conclusion Integrating genomics and metabolomics across independent cohorts enables the discovery of cause-specific biomarkers for the occurrence and progression of CKD in the Korean population.

OBJECTIVES: The coronavirus disease 2019 (COVID-19) pandemic had direct effects on population health through infection and morbidity, as well as indirect effects on population mental health. We estimated the network structure of post-traumatic stress symptoms (PTSS) and depressive symptoms throughout the pandemic in Korea and aimed to identify the most central and bridging symptoms. METHODS: Participants aged 30–64 years completed mental health surveys across 3 phases of the COVID-19 pandemic: March 2020 (n=1,925), February–March 2021 (n=1,754), and December 2021–January 2022 (n=1,595). Using PTSS and depressive symptom data, we conducted network analyses, and the primary measures of symptom importance (centrality) were expected influence and bridge expected influence. RESULTS: In the comorbidity network, although the most central symptoms fluctuated over the course of the pandemic, sleep problems were consistently identified as the most influential bridge symptoms throughout. The symptom network structure differed between the subacute and chronic phases of the pandemic. CONCLUSIONS We found evidence of changes in the network structure of PTSS and depressive symptoms, even as sleep problems retained a consistent role as a bridging symptom. Although overall network structures varied across phases of the pandemic, the bridging role of sleep-related symptoms remained consistently strong, suggesting that sleep problems may represent a general and enduring mechanism underlying PTSS–depression comorbidity. During future pandemics, prompt screening for sleep problems may help prevent the development of comorbidity between PTSS and depressive symptoms.

Objective: To evaluate the adherence of large language model (LLM)-based healthcare research to the Minimum Reporting Items for Clear Evaluation of Accuracy Reports of Large Language Models in Healthcare (MI-CLEAR-LLM) checklist, a framework designed to enhance the transparency and reproducibility of studies on the accuracy of LLMs for medical applications. Materials and Methods: A systematic PubMed search was conducted to identify articles on LLM performance published in high-ranking clinical medicine journals (the top 10% in each of the 59 specialties according to the 2023 Journal Impact Factor) from November 30, 2022, through June 25, 2024. Data on the six MI-CLEAR-LLM checklist items: 1) identification and specification of the LLM used, 2) stochasticity handling, 3) prompt wording and syntax, 4) prompt structuring, 5) prompt testing and optimization, and 6) independence of the test data—were independently extracted by two reviewers, and adherence was calculated for each item. Results: Of 159 studies, 100% (159/159) reported the name of the LLM, 96.9% (154/159) reported the version, and 91.8% (146/159) reported the manufacturer. However, only 54.1% (86/159) reported the training data cutoff date, 6.3% (10/159) documented access to web-based information, and 50.9% (81/159) provided the date of the query attempts. Clear documentation regarding stochasticity management was provided in 15.1% (24/159) of the studies. Regarding prompt details, 49.1% (78/159) provided exact prompt wording and syntax but only 34.0% (54/159) documented prompt-structuring practices. While 46.5% (74/159) of the studies detailed prompt testing, only 15.7% (25/159) explained the rationale for specific word choices. Test data independence was reported for only 13.2% (21/159) of the studies, and 56.6% (43/76) provided URLs for internet-sourced test data. Conclusion Although basic LLM identification details were relatively well reported, other key aspects, including stochasticity, prompts, and test data, were frequently underreported. Enhancing adherence to the MI-CLEAR-LLM checklist will allow LLM research to achieve greater transparency and will foster more credible and reliable future studies.

Background: Respiratory infections play a major role in acute exacerbation of chronic obstructive pulmonary disease (AECOPD). This study assessed the prevalence of bacterial and viral pathogens and their clinical impact on patients with AECOPD. Methods: This retrospective study included 1,186 patients diagnosed with AECOPD at 28 hospitals in South Korea between 2015 and 2018. We evaluated the identification rates of pathogens, basic patient characteristics, clinical features, and the factors associated with infections by potentially drug-resistant (PDR) pathogens using various microbiological tests. Results: Bacteria, viruses, and both were detected in 262 (22.1%), 265 (22.5%), and 129 (10.9%) of patients, respectively. The most common pathogens included Pseudomonas aeruginosa (17.8%), Mycoplasma pneumoniae (11.2%), Streptococcus pneumoniae (9.0%), influenza A virus (19.0%), rhinovirus (15.8%), and respiratory syncytial virus (6.4%). Notably, a history of pulmonary tuberculosis (odds ratio [OR], 1.66; p=0.046), bronchiectasis (OR, 1.99; p=0.032), and the use of a triple inhaler regimen within the past 6 months (OR, 2.04; p=0.005) were identified as significant factors associated with infection by PDR pathogens. Moreover, patients infected with PDR pathogens exhibited extended hospital stays (15.9 days vs. 12.4 days, p=0.018) and higher intensive care unit admission rates (15.9% vs. 9.5%, p=0.030). Conclusion This study demonstrates that a variety of pathogens are involved in episodes of AECOPD. Nevertheless, additional research is required to confirm their role in the onset and progression of AECOPD.

Objective: To evaluate the adherence of large language model (LLM)-based healthcare research to the Minimum Reporting Items for Clear Evaluation of Accuracy Reports of Large Language Models in Healthcare (MI-CLEAR-LLM) checklist, a framework designed to enhance the transparency and reproducibility of studies on the accuracy of LLMs for medical applications. Materials and Methods: A systematic PubMed search was conducted to identify articles on LLM performance published in high-ranking clinical medicine journals (the top 10% in each of the 59 specialties according to the 2023 Journal Impact Factor) from November 30, 2022, through June 25, 2024. Data on the six MI-CLEAR-LLM checklist items: 1) identification and specification of the LLM used, 2) stochasticity handling, 3) prompt wording and syntax, 4) prompt structuring, 5) prompt testing and optimization, and 6) independence of the test data—were independently extracted by two reviewers, and adherence was calculated for each item. Results: Of 159 studies, 100% (159/159) reported the name of the LLM, 96.9% (154/159) reported the version, and 91.8% (146/159) reported the manufacturer. However, only 54.1% (86/159) reported the training data cutoff date, 6.3% (10/159) documented access to web-based information, and 50.9% (81/159) provided the date of the query attempts. Clear documentation regarding stochasticity management was provided in 15.1% (24/159) of the studies. Regarding prompt details, 49.1% (78/159) provided exact prompt wording and syntax but only 34.0% (54/159) documented prompt-structuring practices. While 46.5% (74/159) of the studies detailed prompt testing, only 15.7% (25/159) explained the rationale for specific word choices. Test data independence was reported for only 13.2% (21/159) of the studies, and 56.6% (43/76) provided URLs for internet-sourced test data. Conclusion Although basic LLM identification details were relatively well reported, other key aspects, including stochasticity, prompts, and test data, were frequently underreported. Enhancing adherence to the MI-CLEAR-LLM checklist will allow LLM research to achieve greater transparency and will foster more credible and reliable future studies.

Background: Respiratory infections play a major role in acute exacerbation of chronic obstructive pulmonary disease (AECOPD). This study assessed the prevalence of bacterial and viral pathogens and their clinical impact on patients with AECOPD. Methods: This retrospective study included 1,186 patients diagnosed with AECOPD at 28 hospitals in South Korea between 2015 and 2018. We evaluated the identification rates of pathogens, basic patient characteristics, clinical features, and the factors associated with infections by potentially drug-resistant (PDR) pathogens using various microbiological tests. Results: Bacteria, viruses, and both were detected in 262 (22.1%), 265 (22.5%), and 129 (10.9%) of patients, respectively. The most common pathogens included Pseudomonas aeruginosa (17.8%), Mycoplasma pneumoniae (11.2%), Streptococcus pneumoniae (9.0%), influenza A virus (19.0%), rhinovirus (15.8%), and respiratory syncytial virus (6.4%). Notably, a history of pulmonary tuberculosis (odds ratio [OR], 1.66; p=0.046), bronchiectasis (OR, 1.99; p=0.032), and the use of a triple inhaler regimen within the past 6 months (OR, 2.04; p=0.005) were identified as significant factors associated with infection by PDR pathogens. Moreover, patients infected with PDR pathogens exhibited extended hospital stays (15.9 days vs. 12.4 days, p=0.018) and higher intensive care unit admission rates (15.9% vs. 9.5%, p=0.030). Conclusion This study demonstrates that a variety of pathogens are involved in episodes of AECOPD. Nevertheless, additional research is required to confirm their role in the onset and progression of AECOPD.