Objectives: Large language models (LLMs) demonstrate remarkable potential in healthcare communication. However, whether they can process complex, high-volume medical information, such as stroke-related content, remains insufficiently validated. This study aimed to evaluate the natural language processing capabilities of LLMs in handling such content and to develop an evaluation instrument. Methods: A survey compared educational materials generated by two LLMs (ChatGPT 4.0 and Claude 3) with neurologist-authored content on stroke. The materials were based on two clinical scenarios representing distinct stroke etiologies: cardioembolism and large-artery atherosclerosis. They were evaluated in terms of accuracy, legality, ethics, comprehensiveness, and information delivery. Scores for comprehensiveness and information delivery were compared according to participants’ agreement with the use of LLMs in healthcare. Results: ChatGPT received the highest scores across all domains, except for legality in Scenario 2. In Scenario 1, the ranking for accuracy and summarization of clinical information was, from highest to lowest, ChatGPT, Claude, and the neurologist (η2 = 0.140, p < 0.001; η2 = 0.175, p < 0.001). The same hierarchy was observed in Scenario 2 for accuracy (η2 = 0.077, p < 0.001) and summarization (η2 = 0.194, p < 0.001). Participants who agreed with the use of LLMs in healthcare assigned higher scores for the comprehensiveness (Scenario 1, p = 0.005; Scenario 2, p = 0.007) and information delivery (Scenario 1, p = 0.003; Scenario 2, p = 0.026) of ChatGPT-generated materials than participants who did not agree. Conclusions LLMs demonstrated adequate capability to convey complex content, such as stroke-related information, in an accessible and understandable manner for non-experts.

Neuropathic pain markedly impairs quality of life and imposes a substantial socioeconomic burden, while available treatments often provide only partial relief and are limited by safety concerns. The Neuropathic Pain Special Interest Group of the International Association for the Study of Pain (NeuPSIG-IASP) first published pharmacologic recommendations in 2007, followed by a major update in 2015 and a new guideline in 2025. This narrative review specifically compares the 2015 and 2025 NeuPSIG-IASP guidelines, outlining key methodological changes and therapeutic shifts. The 2025 guideline is based on a larger, more rigorous meta-analysis, maintains α2δ-ligands (adds mirogabalin), serotonin-noradrenaline reuptake inhibitors, and tricyclic antidepressants as first-line drugs, downgrades tramadol into the opioid third-line group. It also introduces high-frequency motor-cortex repetitive transcranial magnetic stimulation as a weakly recommended third-line option and discusses implications for Korean clinical practice.

Flow cytometric lymphocyte subset analysis (FCLSA) is essential for assessing immune status across various diseases and clinical settings. We surveyed current clinical laboratory practices related to FCLSA to establish a baseline reference for future standardization in Korea. Nine university hospitals actively performing FCLSA responded to the 22-question survey, which covered seven categories of laboratory practice. These hospitals used commercial reagent antibody kits from either Beckton Dickinson Biosciences (N = 4) or Beckman Coulter Diagnostics (N = 5). Most hospitals performed daily instrument setup and scheduled maintenance every 2–6 months. Two levels of commercial quality control materials were routinely used each day. Sample and reagent antibody volumes varied across hospitals, even when the same reagent kit was used. Acquired cell counts ranged from 5 × 10 3 to 5 × 10 4 cells, with two hospitals adjusting counts based on the cell type analyzed. Most laboratories reported percentages and general opinions; some additionally reported white blood cell and lymphocyte counts, along with lymphocyte percentages. This is the first comprehensive survey on the clinical laboratory practice of FCLSA in Korea.Standardization of FCLSA should be accelerated to ensure reliable and reproducible results.

Background: Current reference intervals for lymphocyte subpopulations are primarily based on Western populations, with limited data available for Korean children, particularly for extended subsets. We determined absolute cell counts and percentages of lymphocyte subpopulations in Korean children, according to age and sex. Methods: Samples from 92 children—stratified into two age groups, groups 1 (5–9 yrs) and 2 (10–17 yrs)—were obtained. Immunophenotyping was performed via flow cytometry using the Primary Immunodeficiency Orientation Tube (PIDOT) panel, primarily classifying the cells into T, B, and natural killer cell populations. T lymphocytes were divided into CD4+ , CD8+ , and CD4– CD8– subsets; T and B cells were further subdivided according to their maturation stage. Results: Children in group 1 exhibited higher absolute counts of total B cells, unswitched memory B cells/plasma cells, total T cells, CD4+ naïve cells, and TCRγδ+ T cells than those in group 2. In contrast, Group 2 children showed higher absolute counts of CD4+ effector memory (EM) T cells. Males had higher absolute counts of total B cells, particularly pregerminal center B cells, CD4+ EM cells, and CD8+ terminally differentiated T cells, whereas females showed higher proportions of CD4+ , CD4+ naïve, and CD8 + central memory/transitional memory T cells. Conclusions To the best of our knowledge, this study is the first to establish reference values for extended lymphocyte subsets in Korean children using the PIDOT panel. Age, sex, and laboratory-related factors influenced lymphocyte subset distributions. These findings may serve as reference data for immune disorders and immunotherapy in pediatric populations.

In forensic medicine, age estimation commonly involves assessing age-related changes in teeth and skeletal structures. Vertebral morphological alterations, such as osteophyte formation, serve as age indicators. Recent studies using deep-learning techniques, such as neural networks, for age estimation from radiographic images have been conducted, reporting significantly higher accuracy than previous studies. This study aimed to estimate age using neural network-based deep-learning techniques applied to computed tomography (CT) cross-sectional images of the spine and evaluate its feasibility. Postmortem CT scans of 214 cadavers with varying decomposition levels were used. Coronal and sagittal cross-sectional images penetrating the center of each vertebral body were extracted for the 11th and 12th thoracic vertebrae and the first to fifth lumbar vertebrae. Using these images, along with the chronological ages of deceased individuals, an age estimation model was developed through regression analysis in PyTorch, employing a convolutional neural networks architecture with five-fold cross-validation. The model achieved a mean absolute error of 5.385 years, root mean squared error of 7.029 years, and coefficient of determination of 0.793. Although the sample size was relatively small, the results suggested the potential applicability of vertebral imagingbased age estimation in the Korean population. Further research using a larger dataset may improve the accuracy and reliability of the model.

Purpose: While MRI-based tumor regression grade (mrTRG) has shown promise in evaluating pathologic response to concurrent chemoradiotherapy (CCRT) in rectal cancer, its ability to predict pathologic complete response remains limited.This study aimed to enhance mrTRG’s diagnostic performance in predicting ypT0–1N0 status, a key factor in considering non-radical management after CCRT for locally advanced rectal cancer (LARC). Methods: This retrospective study included 430 patients with LARC who underwent radical resection following CCRT at a single referral hospital between April 2018 and September 2024. Multivariable logistic regression was used to identify predictive factors associated with achieving ypT0–1N0 status. The diagnostic performances of mrTRG1–2 alone and in combination with other factors were assessed by comparing sensitivity, specificity, positive-predictive value (PPV), negative-predictive value, and area under the curve (AUC). Results: Ninety-three patients (21.6%) achieved ypT0–1N0. In the multivariable analysis, fungating type, cT1–2, and mrTRG1–2 were independent predictors for ypT0–1N0. Integrating mrTRG with gross tumor type yielded the highest AUC of 0.689 among the combined models. For predicting ypT0–1N0, the combination of mrTRG and gross tumor type improved PPV (79.2% vs. 41.5% for mrTRG alone) while also demonstrating enhanced sensitivity compared with ycT0–1N0, the conventional MRI-based predictor (40.9% vs. 22.6%). Conclusion This study demonstrated that combining mrTRG and gross tumor type improved the PPV of mrTRG in predicting ypT0–1N0 after CCRT in LARC. Further studies are warranted to validate the role of gross tumor type in refining predictive systems for selecting candidates for non-radical treatment.

Buttock augmentation is an increasingly popular cosmetic procedure designed to enhance buttock contour, size, and shape. However, the safety profile of this procedure remains insufficiently established, and it carries risks of complications, including foreign body reactions and infections. These complications may be exacerbated by filler migration, resulting in large soft-tissue cavities that resemble pseudocysts. In this study, we describe two patients who developed severe complications following massive filler injections to the buttocks. A 56-year-old female patient presented with a 6×5 cm soft-tissue defect associated with an extensive underlying dead space, sinus tract formation, and a large pocket extending across the buttock. Additionally, a 50-year-old female patient developed diffuse cellulitis and multiple abscesses secondary to migration of an infected filler-related pseudocyst. Both patients underwent successful en bloc capsulectomy, resulting in marked clinical improvement without recurrence or postoperative complications. These cases underscore the serious complications associated with large-volume filler injections and highlight the importance of comprehensive surgical management in addressing late-stage adverse outcomes.

Purpose: Accumulating evidence has clarified that gut dysbiosis is involved in lung cancer development and progression. Although the relationship between tumors and gut microbiota has been extensively studied using clinical samples, no studies have examined the association between mutant epidermal growth factor receptor (EGFR)–induced lung carcinogenesis and dysbiosis in gut microbiota. Therefore, we investigated the gut microbiota profiles in stool samples from human lung-specific conditional EGFR-mutant transgenic mice during lung tumor carcinogenesis. Materials and Methods: Stool samples were collected before tamoxifen treatment (V1) and at each time point following mutant EGFR expression in lung tissue (V2) and lung tumor appearance (V3). Fecal 16S rRNA taxonomy was analyzed to assess microbial diversity, composition, and dynamic changes at each time point. Results: We found that microbiota richness and diversity were significantly elevated when tumors developed and grew in the lung. Phylogenetic analysis of the microbial community revealed that Lachnospiraceae, Ruminococcaceae, Porphyromonadaceae, Rhodospirillaceae, Odoribacteraceae, and Desulfovibrionaceae showed a significant increase at the V3 stage compared to the V1 stage at the family level. In contrast, Lactobacillaceae, Bacteroidaceae, Muribaculaceae, Coriobacteriaceae, and Rikenellaceae significantly decreased at the V3 stage compared to the V1 stage. Furthermore, Lactobacillus species, also known as short chain fatty acid-producing bacteria, were relatively abundant at the V1 stage but were depleted with the occurrence of lung tumors at the V3 stage. Conclusion Changes in gut microbiota, such as Lactobacillus species, may be a predictive factor for the emergence and progression of tumors in an animal model of lung adenocarcinoma induced by mutant EGFR.

The sixth edition of the Korean Medication Algorithm Project for Depressive Disorder (KMAP-DD) was published in 2025. This review compared KMAP-DD 2025 with four major international clinical practice guidelines: Canadian Network for Mood and Anxiety Treatments Clinical Guidelines for the Management of Major Depressive Disorders, National Institute for Health and Care Excellence Depression Guideline, Royal Australian and New Zealand College of Psychiatrists Clinical Practice Guidelines for Mood Disorders, and British Association for Psychopharmacology Guideline. While KMAP-DD is based on expert consensus, and others on evidence-based methods, overall treatment strategies for depressive episodes were fairly consistent. Especially, KMAP-DD 2025 offers more structured recommendations in areas lacking strong evidence, such as premenstrual dysphoric disorder, perinatal depression, and depression with medical comorbidities. KMAP-DD 2025 also reflected Korean clinical practice patterns emphasizing rapid symptom relief and early use of combination strategies. Despite limitations as a consensus-based guideline, KMAP-DD 2025 complements evidence-based approaches and provides practical, situation-specific guidance for real-world clinical decision-making in Korea.