Background: This study aimed to evaluate the applicability of deep learning technology to thyroid ultrasound images for classification of thyroid nodules. Methods: This retrospective analysis included ultrasound images of patients with thyroid nodules investigated by fine-needle aspiration at the thyroid clinic of a single center from April 2010 to September 2012. Thyroid nodules with cytopathologic results of Bethesda category V (suspicious for malignancy) or VI (malignant) were defined as thyroid cancer. Multiple deep learning algorithms based on convolutional neural networks (CNNs) —ResNet, DenseNet, and EfficientNet—were utilized, and Siamese neural networks facilitated multi-view analysis of paired transverse and longitudinal ultrasound images. Results: Among 1,048 analyzed thyroid nodules from 943 patients, 306 (29%) were identified as thyroid cancer. In a subgroup analysis of transverse and longitudinal images, longitudinal images showed superior prediction ability. Multi-view modeling, based on paired transverse and longitudinal images, significantly improved the model performance; with an accuracy of 0.82 (95% confidence intervals [CI], 0.80 to 0.86) with ResNet50, 0.83 (95% CI, 0.83 to 0.88) with DenseNet201, and 0.81 (95% CI, 0.79 to 0.84) with EfficientNetv2_ s. Training with high-resolution images obtained using the latest equipment tended to improve model performance in association with increased sensitivity. Conclusion CNN algorithms applied to ultrasound images demonstrated substantial accuracy in thyroid nodule classification, indicating their potential as valuable tools for diagnosing thyroid cancer. However, in real-world clinical settings, it is important to aware that model performance may vary depending on the quality of images acquired by different physicians and imaging devices.

Background: This study aimed to evaluate the applicability of deep learning technology to thyroid ultrasound images for classification of thyroid nodules. Methods: This retrospective analysis included ultrasound images of patients with thyroid nodules investigated by fine-needle aspiration at the thyroid clinic of a single center from April 2010 to September 2012. Thyroid nodules with cytopathologic results of Bethesda category V (suspicious for malignancy) or VI (malignant) were defined as thyroid cancer. Multiple deep learning algorithms based on convolutional neural networks (CNNs) —ResNet, DenseNet, and EfficientNet—were utilized, and Siamese neural networks facilitated multi-view analysis of paired transverse and longitudinal ultrasound images. Results: Among 1,048 analyzed thyroid nodules from 943 patients, 306 (29%) were identified as thyroid cancer. In a subgroup analysis of transverse and longitudinal images, longitudinal images showed superior prediction ability. Multi-view modeling, based on paired transverse and longitudinal images, significantly improved the model performance; with an accuracy of 0.82 (95% confidence intervals [CI], 0.80 to 0.86) with ResNet50, 0.83 (95% CI, 0.83 to 0.88) with DenseNet201, and 0.81 (95% CI, 0.79 to 0.84) with EfficientNetv2_ s. Training with high-resolution images obtained using the latest equipment tended to improve model performance in association with increased sensitivity. Conclusion CNN algorithms applied to ultrasound images demonstrated substantial accuracy in thyroid nodule classification, indicating their potential as valuable tools for diagnosing thyroid cancer. However, in real-world clinical settings, it is important to aware that model performance may vary depending on the quality of images acquired by different physicians and imaging devices.

Background: This study aimed to evaluate the applicability of deep learning technology to thyroid ultrasound images for classification of thyroid nodules. Methods: This retrospective analysis included ultrasound images of patients with thyroid nodules investigated by fine-needle aspiration at the thyroid clinic of a single center from April 2010 to September 2012. Thyroid nodules with cytopathologic results of Bethesda category V (suspicious for malignancy) or VI (malignant) were defined as thyroid cancer. Multiple deep learning algorithms based on convolutional neural networks (CNNs) —ResNet, DenseNet, and EfficientNet—were utilized, and Siamese neural networks facilitated multi-view analysis of paired transverse and longitudinal ultrasound images. Results: Among 1,048 analyzed thyroid nodules from 943 patients, 306 (29%) were identified as thyroid cancer. In a subgroup analysis of transverse and longitudinal images, longitudinal images showed superior prediction ability. Multi-view modeling, based on paired transverse and longitudinal images, significantly improved the model performance; with an accuracy of 0.82 (95% confidence intervals [CI], 0.80 to 0.86) with ResNet50, 0.83 (95% CI, 0.83 to 0.88) with DenseNet201, and 0.81 (95% CI, 0.79 to 0.84) with EfficientNetv2_ s. Training with high-resolution images obtained using the latest equipment tended to improve model performance in association with increased sensitivity. Conclusion CNN algorithms applied to ultrasound images demonstrated substantial accuracy in thyroid nodule classification, indicating their potential as valuable tools for diagnosing thyroid cancer. However, in real-world clinical settings, it is important to aware that model performance may vary depending on the quality of images acquired by different physicians and imaging devices.

Background: This study aimed to evaluate the applicability of deep learning technology to thyroid ultrasound images for classification of thyroid nodules. Methods: This retrospective analysis included ultrasound images of patients with thyroid nodules investigated by fine-needle aspiration at the thyroid clinic of a single center from April 2010 to September 2012. Thyroid nodules with cytopathologic results of Bethesda category V (suspicious for malignancy) or VI (malignant) were defined as thyroid cancer. Multiple deep learning algorithms based on convolutional neural networks (CNNs) —ResNet, DenseNet, and EfficientNet—were utilized, and Siamese neural networks facilitated multi-view analysis of paired transverse and longitudinal ultrasound images. Results: Among 1,048 analyzed thyroid nodules from 943 patients, 306 (29%) were identified as thyroid cancer. In a subgroup analysis of transverse and longitudinal images, longitudinal images showed superior prediction ability. Multi-view modeling, based on paired transverse and longitudinal images, significantly improved the model performance; with an accuracy of 0.82 (95% confidence intervals [CI], 0.80 to 0.86) with ResNet50, 0.83 (95% CI, 0.83 to 0.88) with DenseNet201, and 0.81 (95% CI, 0.79 to 0.84) with EfficientNetv2_ s. Training with high-resolution images obtained using the latest equipment tended to improve model performance in association with increased sensitivity. Conclusion CNN algorithms applied to ultrasound images demonstrated substantial accuracy in thyroid nodule classification, indicating their potential as valuable tools for diagnosing thyroid cancer. However, in real-world clinical settings, it is important to aware that model performance may vary depending on the quality of images acquired by different physicians and imaging devices.

This study examined the colostrum nutritive composition, immunoglobulin (Ig), and microbial community in Holstein and Jersey dairy cows according to the time after calving. The experiment used seven Holstein and three Jersey dairy cows. Colostrum was collected immediately after calf calving, 12, and 24 hours, and stored at −80°C until analysis. An analysis of the nutritive composition in colostrum was performed using LactoScop. The immune indicators were analyzed using an ELISA Kit, and the microbial community was assessed using a Macrogen Inc. The protein level was high in all colostrum samples from Holstein dairy cows compared with Jersey dairy cows, but there was no significant difference according to the time after calving. Immune index analysis revealed high IgG and IgA concentrations in the colostrum of Holstein cows immediately after calving and 12 and 24 hours after calving, but the differences were not significant. The microbial community at the genus level revealed Staphylococcus to be predominant at a high rate in the colostrum of Holstein dairy cows and Enterococcus in Jersey dairy cows 12 hours after calving. Pseudomonas was predominant at a high rate in the colostrum of Jersey lactating cows immediately and 12 hours after calving. Chryseobacterium was predominant at a high rate in Holstein dairy cows 12 and 24 hours after calving. In conclusion, these results are expected to be used as research data on the correlation between quality, immunity, and microbial community in the colostrum. In the future, beneficial microorganisms in the colostrum of domestic dairy cows can be used to improve the growth and immunity of Holstein and Jersey calves and assist in research related to postbiotics industrialization.

BACKGROUND/OBJECTIVES: High levels of plasma low-density lipoprotein (LDL) cholesterol are an important determinant of atherosclerotic lesion formation. The disruption of cholesterol efflux or reverse cholesterol transport (RCT) in peripheral tissues and macrophages may promote atherogenesis. The aim of the current study was to examine whether bioactive ellagic acid, a functional food component, improved RCT functionality and high-density lipoprotein (HDL) function in diet-induced atherogenesis of apolipoproteins E (apoE) knockout (KO) mice.MATERIALS/METHODS: Wild type mice and apoE KO mice were fed a high-cholesterol Paigen diet for 10 weeks to induce hypercholesterolemia and atherosclerosis, and concomitantly received 10 mg/kg ellagic acid via gavage. RESULTS: Supplying ellagic acid enhanced induction of apoE and ATP-binding cassette (ABC) transporter G1 in oxidized LDL-exposed macrophages, facilitating cholesterol efflux associated with RCT. Oral administration of ellagic acid to apoE KO mice fed on Paigen diet improved hypercholesterolemia with reduced atherogenic index. This compound enhanced the expression of ABC transporters in peritoneal macrophages isolated from apoE KO mice fed on Paigen diet, indicating increased cholesterol efflux. Plasma levels of cholesterol ester transport protein and phospholipid transport protein involved in RCT were elevated in mice lack of apoE gene, which was substantially reduced by supplementing ellagic acid to Paigen diet-fed mice. In addition, ellagic acid attenuated hepatic lipid accumulation in apoE KO mice, evidenced by staining of hematoxylin and eosin and oil red O. Furthermore, the supplementation of 10 mg/kg ellagic acid favorably influenced the transcriptional levels of hepatic LDL receptor and scavenger receptor-B1 in Paigen diet-fed apoE KO mice. CONCLUSION Ellagic acid may be an athero-protective dietary compound encumbering diet-induced atherogenesis though improving the RCT functionality.

Purpose: Acetaminophen overdose accounts for the largest proportion of pediatric poisoning reported in South Korea. We investigated the characteristics and prognosis of pediatric acetaminophen overdose. Methods: We retrospectively reviewed medical records of patients younger than 19 years with acetaminophen overdose who visited the emergency department (ED) of Pusan National University Children’s Hospital from January 2009 through December 2022. We investigated age, sex, dose, clinical findings, and treatment, and compared the differences in the variables according to the intentionality of ingestion and the presence of hepatotoxicity. Results: Among the 132,691 pediatric patients who visited the ED during the period, 555 did for drug intoxication, of whom 51 with acetaminophen overdose were included in this study. The overdose was intentional in 43 patients (the intentional group; median age, 15.4 years [interquartile range, 13.9-16.8]) and accidental in 8 patients (the accidental group; 7.7 years [2.3-15.4]). The mean ingested dose was higher in the intentional group than in the accidental group (170.3 ± 129.0 vs. 105.3 ± 37.5 mg/kg; P = 0.016). Hospitalization tended to be implemented more frequently in the intentional group (53.5% vs. 12.5%; P = 0.081). All patients with hepatotoxicity (n = 10) belonged to the same group, and showed a higher or longer median age (15.9 [14.6-17.1] vs. 13.8 years [13.7-16.4]; P = 0.030), mean ingested dose (13,230.0 ± 10,544.8 vs. 7,654.0 ± 6,752.3 mg; P = 0.043), mean time from ingestion to arrival at the ED (22.6 ± 23.4 vs. 5.5 ± 6.4 hours; P = 0.048), and frequency of hospitalization (80.0% vs. 39.0%; P = 0.048). Conclusion It is necessary to track the occurrence of hepatotoxicity in patients who intentionally ingest acetaminophen.

Background/Aims: Long-term use of acid suppressants such as proton pump inhibitors (PPIs) and histamine 2 receptor antagonist (H2RA) has been associated with the risk of osteoporotic fracture. Acid suppressants and muco-protective agents (MPAs) are often used together as anti-ulcer agents. We evaluated the association between the risk of osteoporotic fracture and the combined use of these anti-peptic agents. Methods: A population-based case-control study was conducted by analyzing the Korean National Health Insurance Data from 2014 to 2020. Patients who had been prescribed anti-peptic agents, such as PPI, H2RA, or MPA, were included. Considering the incidence of osteoporotic fractures, the case group (n = 14,704) and control group (n = 58,816) were classified by 1:4 matching based on age and sex. Results: The use of all types of anti-peptic agents was associated with an increased risk of osteoporotic fractures (PPI: hazard osteoratio [HR], 1.31; H2RA: HR, 1.44; and MPA: HR, 1.33; all p < 0.001). Compared to PPI alone, the combined use of “PPI and H2RA” (HR, 1.58; p = 0.010) as well as “PPI, H2RA, and MPA” (HR, 1.71; p = 0.001) was associated with an increased risk of osteoporotic fracture. However, compared with PPI alone, “MPA and PPI or H2RA” was not associated with an increased risk of osteoporotic fracture. Conclusions This study found that the combined use of “PPI and H2RA” was associated with a higher risk of osteoporotic fractures. In cases where deemed necessary, the physicians may initially consider prescribing the combination use of MPA.

Background/Aims: Mucoprotective agents, such as eupatilin, are often prescribed to prevent gastrointestinal (GI) bleeding in addition to an acid suppressant despite the absence of a large-scale study. We evaluated the additional effect of eupatilin on the prevention of GI bleeding in both the upper and lower GI tract in concomitant aspirin and acid suppressant users using the nationwide database of national claims data from the Korean National Health Insurance Service (NHIS). Methods: An aspirin cohort was constructed using the NHIS claims data from 2013 to 2020. Patients who manifested with hematemesis, melena, or hematochezia were considered to have GI bleeding. A Cox proportional hazards regression model was used to determine the risk factors for GI bleeding associated with the concomitant use of GI drugs and other covariates among aspirin users. Results: Overall, a total of 432,208 aspirin users were included. The concurrent use of an acid suppressant and eupatilin (hazard ratio [HR] = 0.85, p = 0.016, vs. acid suppressant only) was a statistically significant preventive factor for GI bleeding. Moreover, a more than 3-month duration (HR = 0.88, p = 0.030) of acid suppressant and eupatilin prescription (vs. acid suppressant only) was a statistically significant preventive factor for GI bleeding. Conclusions Eupatilin administration for ≥ 3 months showed additional preventive effect on GI bleeding in concomitant aspirin and acid suppressant users. Thus, cotreatment with eupatilin with a duration of 3 months or longer is recommended for reducing GI bleeding among aspirin plus acid suppressant users.

Objective: To determine the clinical impact of the baseline sagittal imbalance severity in patients with adult spinal deformity (ASD). Methods: We retrospectively reviewed patients who underwent ≥ 5-level fusion including the pelvis, for ASD with a ≥ 2-year follow-up. Using the Scoliosis Research Society-Schwab classification system, patients were classified into 3 groups according to the severity of the preoperative sagittal imbalance: mild, moderate, and severe. Postoperative clinical and radiographic results were compared among the 3 groups. Results: A total of 259 patients were finally included. There were 42, 62, and 155 patients in the mild, moderate, and severe groups, respectively. The perioperative surgical burden was greatest in the severe group. Postoperatively, this group also showed the largest pelvic incidence minus lumbar lordosis mismatch, suggesting a tendency towards undercorrection. No statistically significant differences were observed in proximal junctional kyphosis, proximal junctional failure, or rod fractures among the groups. Visual analogue scale for back pain and Scoliosis Research Society-22 scores were similar across groups. However, severe group’s last follow-up Oswestry Disability Index (ODI) scores significantly lower than those of the severe group. Conclusion Patients with severe sagittal imbalance were treated with more invasive surgical methods along with increased the perioperative surgical burden. All patients exhibited significant radiological and clinical improvements after surgery. However, regarding ODI, the severe group demonstrated slightly worse clinical outcomes than the other groups, probably due to relatively higher proportion of undercorrection. Therefore, more rigorous correction is necessary to achieve optimal sagittal alignment specifically in patients with severe baseline sagittal imbalance.