Background: Advancing orthopedic care through large language models requires both multimodal processing capabilities for medical images and open-source deployment options for secure in-house operations, yet these remain underexplored in current literature. This study aims to benchmark open-source vision-language models (VLMs) against orthopedic residents using the Orthopedic In-Training Examination (OITE), assess domain-specific performance across orthopedic subspecialties, and investigate the relationship between model parameter size and performance. Methods: Six open-source VLMs of varying sizes (Alibaba Qwen2.5-VL-72B-Instruct, Alibaba Qwen2.5-VL-32B-Instruct, Alibaba Qwen2.5-VL-7B-Instruct, Alibaba Qwen2.5-VL-3B-Instruct, Meta Llama-3.2-90B-Vision-Instruct, Meta Llama-3.2-11B-Vision-Instruct) were evaluated using the 2023 OITE (210 questions; 111 with images). Model performance was compared to resident scores from the 2023 OITE technical report. Pearson correlation coefficient was used to assess the association between model size and performance. Results: The 2 largest open-source models, Qwen2.5-VL-72B and Llama-3.2-90B, demonstrated performance levels comparable to those of second-year orthopedic residents on the OITE examination. A mid-sized model, Qwen-32B, slightly outscored first-year residents. In contrast, small-sized models (under 11 billion parameters) performed worse than first-year residents. Qwen2.5-VL-72B performed best in foot & ankle and sports medicine topics, while Llama-3.2-90B was strongest in basic science and hand & wrist.All models had the most difficulty with spine and pediatric questions. Overall, model accuracy increased steadily with model size up to 72 billion parameters, but larger sizes showed little additional improvement. Conclusions Smaller models offer reduced accuracy in exchange for lower hardware requirements. Spine and pediatric domains remain consistently areas of underperformance across all models. Model selection should be based on domain-specific benchmark results to balance clinical needs with hardware limitations. While promising, open-source VLMs currently require further refinement and validation before they can be reliably applied in clinical or educational settings.

Objective: Depression involves mood-related behavioral changes typically monitored through subjective reports, which are limited by recall bias and low temporal resolution. Digital mental health tools offer objective, continuous monitoring, but prior studies have focused on outpatients subject to environmental variability. In this preliminary feasibility study, we examined psychiatric inpatients in a controlled setting to assess associations between behavioral regularity and depression severity, highlighting the clinical potential of digital phenotyping. Methods: Thirty-five adults from a closed psychiatric ward were recruited, and data from 10 inpatients with ≥7 days of valid monitoring were analyzed. Depression severity was assessed weekly using the Hamilton Depression Rating Scale (HAMD) and Dysfunctional Self-focus Attributes Scale, yielding 18 samples. Hourly accelerometer and location data from wearable devices and ward sensors were processed to generate digital phenotypes—interdaily stability (IS), intradaily variability (IV), ratio of IS to IV (ISV), entropy (EN), and normalized entropy (NE)—segmented into daytime and nighttime. Linear mixed models assessed group differences, and correlation and multiple regression examined associations with depression. Results: Patients with asymptomatic/mild depression showed significantly higher IS_day and ISV_day, and lower EN_night, and NE_night (all p<0.05). These four features correlated with HAMD after false discovery rate (all p<0.05) correction. A regression model including IS_day and NE_night explained 60.6% of HAMD variance (p<0.05). Conclusion Digital monitoring provides an objective and continuous method to assess depression severity. By capturing macro- and micro-level movement regularity across day and night in an inpatient environment, this approach offers practical relevance for psychiatric care. However, results should be considered preliminary due to the limited sample size.

Extracellular vesicles are nano-sized vesicles surrounded by lipid bilayers, and all cells release them to the extracellular environment for communication. Extracellular vesicles consist of molecules with various biological activities and can play essential roles as therapeutics, so they attract much attention as next-generation modalities to treat various diseases. As extracellular vesicles are cell-derived nanovesicles, they are favorable to be developed as therapeutics, but they also have limitations. In addition, there are a number of things to consider in terms of manufacturing, quality control, non-clinical studies, and clinical trials during the development of extracellular vesicle-based therapeutics. Meanwhile, as much attention has been paid to the potentials of extracellular vesicles as therapeutics, many biopharmaceutical companies are trying to develop extracellular vesicle-based therapeutics. This review will introduce the advantages and limitations of extracellular vesicles as therapeutics. In addition, it will cover things to consider during developing extracellular vesicle-based therapeutics and development cases of extracellular vesicle-based therapeutics.

Purpose: The prevalence of adolescents with type 2 diabetes mellitus (T2DM) has rapidly increased in Korea over the past few decades with the increase in the number of obese adolescents. The single point insulin sensitivity estimator (SPISE) was recently introduced as a surrogate marker for insulin sensitivity to predict T2DM in adults. We aimed to determine risk factors for T2DM in obese adolescents, including SPISE. Methods: This retrospective study included 104 adolescents diagnosed with T2DM at Korea University Hospital between January 2010 and December 2020. We compared clinical and biochemical parameters and the SPISE of normoglycemic overweight and obese individuals with those of prediabetic and diabetic adolescents to determine risk factors for T2DM. Receiver operating characteristic analysis was performed with the Youden index to determine the cutoff point of SPISE. Results: Frequency of fatty liver and family history of T2DM were significantly higher and SPISE level was significantly lower in patients with T2DM than in normoglycemic overweight/obese and prediabetic adolescents (p<0.01). A family history of T2DM, fatty liver, and SPISE value below the cutoff point (4.49) were identified as significant risk factors for T2DM in multiple logistic regression analysis after controlling for age, sex, and body mass index standard deviation score (p<0.01). Conclusion Family history of T2DM, fatty liver, and low SPISE (<4.49) are risk factors that can independently affect the occurrence of T2DM in obese adolescents. Among these risk factors, SPISE is a promising marker for predicting adolescent T2DM; careful monitoring of these individuals is needed to prevent progression to T2DM.

Purpose: Inhaled nitric oxide (iNO) is a potent selective pulmonary vasodilator and an important treatment for newborn infants with hypoxic respiratory failure due to persistent pulmonary hypertension of the newborn (PPHN). The Ministry of Food and Drug Safety of Korea first approved iNO in 2009 for use as a new drug to treat hypoxic res piratory failure with PPHN. A post-marketing surveillance study (PMSS) was conducted to assess the effectiveness and safety of the iNO treatment. We evaluated the clinical status of the iNO treatment currently available in Korea by using the PMSS data.
Methods: We retrospectively reviewed the PMSS data from 22 hospitals in Korea from October 2014 to September 2018. Altogether, 97 infants were enrolled and divided into early response (ER), delayed response (DR), no response (NR), and death groups according to their response to the iNO treatment.
Results: The ER group included 52 infants (53.6%); DR group, 33 (34.0%); NR group, seven (7.2%); and death group, five (5.2%). The iNO treatment was initiated within 14 days after birth at a concentration of 20 ppm. The median treatment duration was 91.5 hours (69.0 to 134.3) in all the infants. Of the infants, 43 (44.3%) received the treatment for >96 hours. Fifty-one infants (52.6%) needed >20 ppm of iNO, and 10 (10.5%) needed a maximum of 80 ppm of iNO. None of the infants had hypermethemoglobinemia or an alarm report of NO2.
Conclusion Korean neonatologists reported that a large proportion of the infants with PPHN showed improvement in oxygenation after treatment with the approved iNO. These infants received relatively longer durations and higher concentrations of the iNO treatment than the current recommendations, without any side and adverse effect.

No abstract available.
Bone Marrow ; Lymphoma ; Paraproteinemias

No abstract available.
Bone Marrow* ; Lymphoma, Follicular* ; Lymphoma, Mantle-Cell*

Paroxysmal dysarthria and ataxia is characterized by abrupt onset of dysarthria and ataxia of unilateral limbs for seconds. We present a 45-year-old female patient with paroxysmal symptoms of dysarthria, right-sided ataxia, and a persistent symptom of upbeating nystagmus. Her brain fluid attenuated inversion recovery (FLAIR) magnetic resonance imaging indicated potential diagnosis of demyelinating disease. The paroxysmal and persistent symptoms have recovered after phenytoin administration. Sodium channel blocker may play a role in reducing the ephaptic transmission in a demyelinated or re-myelinated lesion.
Ataxia ; Brain ; Demyelinating Diseases ; Diagnosis ; Dysarthria ; Extremities ; Female ; Humans ; Magnetic Resonance Imaging ; Middle Aged ; Phenytoin ; Sodium Channels

OBJECTIVES: Obsessive-compulsive disorder (OCD) is a disabling psychiatric disorder, and more attention is recently paid on the quality of life (QoL) in OCD patients. The Euro-QoL-5D (EQ-5D) is a widely used self-report to calculate a single score which represents ‘health utilities’. The aim of this study was to assess the health-related QoL for patients with OCD using the EQ-5D and to examine the relationship between health-related QoL and symptom severity. METHODS: Seventy-three patients with a Diagnostic and Statistical Manual of Mental Disorders (DSM-5) diagnosis of OCD were recruited from the outpatient clinic in Seoul National University Hospital. Symptom severity was assessed using the Yale-Brown Obsessive-Compulsive Scale (YBOCS), and QoL was assessed with the EQ-5D-5L questionnaire. Using Korean valuation study, a single score of QoL was calculated. RESULTS: Most of the OCD patients were relatively young (< 45 years) with the mean YBOCS total score of 19.33. The mean EQ-5D score was 0.71 and significantly correlated with symptom severity (r = -0.483, p < 0.001). 25% of the EQ-5D score was predicted by the YBOCS total score (b = -0.011, p < 0.001) by regression analysis. CONCLUSIONS: OCD patients suffer from lower health-related QoL and QoL significantly decreased as symptom severity increased. The results of the EQ-5D would enable further studies on QoL comparison across medical disease and mental disorders.
Ambulatory Care Facilities ; Diagnosis ; Diagnostic and Statistical Manual of Mental Disorders ; Humans ; Mental Disorders ; Obsessive-Compulsive Disorder* ; Quality of Life* ; Seoul

OBJECTIVES: Obsessive-compulsive disorder (OCD) is a chronic and disabling psychiatric disorder. The duration of untreated illness (DUI) has been suggested as one of the predictors of clinical course and outcome in various psychiatric disorders. There is increasing evidence that cognitive dysfunction is associated with the prognosis of OCD. The aim of this study was to investigate the influence of DUI on the neurocognitive functions in patients with OCD. METHODS: Sixty-two patients with a DSM-IV diagnosis of OCD from the outpatient clinic were included in this study. We defined the short DUI if the DUI was 2-year or less and the long DUI if it was longer than 2-year. Neurocognitive functions were assessed by visuospatial memory function test and 4 subsets of K-WAIS such as vocabulary, arithmetic, block design and picture arrangement. Differences in neurocognitive functions as well as clinical variables between OCD patients with short DUI and those with long DUI were investigated. Correlation analyses were also performed to determine the correlation between DUI and neurocognitive functions. RESULTS: Compared with the short DUI group, the long DUI group performed worse in the block design test, which measures executive function. The long DUI group also had a higher level of compulsive symptom severity than the short DUI group. However, the DUI was not correlated with neurocognitive functions. CONCLUSIONS: Findings in this preliminary study suggest that the long DUI in patients with OCD is associated with more severe executive dysfunction. Studies with larger samples and longitudinal design are needed to further confirm the prognostic role of the DUI in OCD.
Ambulatory Care Facilities ; Diagnosis ; Diagnostic and Statistical Manual of Mental Disorders ; Executive Function ; Humans ; Memory ; Obsessive-Compulsive Disorder* ; Prognosis ; Vocabulary