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 aimed to elucidate the clinical characteristics of patients diagnosed with toxoplasmosis in Korea. We collected and analyzed the specific research data of 5,917 patients from the Health Insurance Review and Assessment (HIRA; 2007–2020) and 533 electronic medical records (EMRs; 2003–2021) of Korean patients. The HIRA data showed that toxoplasmosis is an endemic disease that occurs constantly in Korea, with a large proportion of patients complaining of ocular symptoms. Of the 533 patients for whom EMR data were available, 54.6% were diagnosed with toxoplasmosis; ocular toxoplasmosis (35.7%), congenital toxoplasmosis (4.7%), cerebral toxoplasmosis (4.1%), pulmonary toxoplasmosis (0.4%), and toxoplasma hepatitis (0.6%), in order of frequency. In ocular cases, 54.4% of the patients had diverse ocular pathologies. Toxoplasmosis in Korea is characterized by a high frequency of ocular symptoms, most patients are adults, and 51.8% of patients with seropositivity were positive for IgG, suggesting prior infection. This study highlights that patients with ocular symptoms are included in the major diagnosis group for acquired toxoplasmosis in Korea.

This study aimed to elucidate the clinical characteristics of patients diagnosed with toxoplasmosis in Korea. We collected and analyzed the specific research data of 5,917 patients from the Health Insurance Review and Assessment (HIRA; 2007–2020) and 533 electronic medical records (EMRs; 2003–2021) of Korean patients. The HIRA data showed that toxoplasmosis is an endemic disease that occurs constantly in Korea, with a large proportion of patients complaining of ocular symptoms. Of the 533 patients for whom EMR data were available, 54.6% were diagnosed with toxoplasmosis; ocular toxoplasmosis (35.7%), congenital toxoplasmosis (4.7%), cerebral toxoplasmosis (4.1%), pulmonary toxoplasmosis (0.4%), and toxoplasma hepatitis (0.6%), in order of frequency. In ocular cases, 54.4% of the patients had diverse ocular pathologies. Toxoplasmosis in Korea is characterized by a high frequency of ocular symptoms, most patients are adults, and 51.8% of patients with seropositivity were positive for IgG, suggesting prior infection. This study highlights that patients with ocular symptoms are included in the major diagnosis group for acquired toxoplasmosis in Korea.

Background: The role of apurinic/apyrimidinic endonuclease 1/redox factor-1 (APE1/Ref-1) in adipose tissue remains poorly understood. This study investigates adipose tissue dysfunction in heterozygous APE1/Ref-1 deficiency (APE1/Ref-1+/-) mice, focusing on changes in adipocyte physiology, oxidative stress, adipokine regulation, and adipose tissue distribution. Methods: APE1/Ref-1 mRNA and protein levels in white adipose tissue (WAT) were measured in APE1/Ref-1+/- mice, compared to their wild-type (APE1/Ref-1+/+) controls. Oxidative stress was assessed by evaluating reactive oxygen species (ROS) levels. Histological and immunohistochemical analyses were conducted to observe adipocyte size and macrophage infiltration of WAT. Adipokine expression was measured, and micro-magnetic resonance imaging (MRI) was used to quantify abdominal fat volumes. Results: APE1/Ref-1+/- mice exhibited significant reductions in APE1/Ref-1 mRNA and protein levels in WAT and liver tissue. These mice also showed elevated ROS levels, suggesting a regulatory role for APE1/Ref-1 in oxidative stress in WAT and liver. Histological and immunohistochemical analyses revealed hypertrophic adipocytes and macrophage infiltration in WAT, while Oil Red O staining demonstrated enhanced ectopic fat deposition in the liver of APE1/Ref-1+/- mice. These mice also displayed altered adipokine expression, with decreased adiponectin and increased leptin levels in the WAT, along with corresponding alterations in plasma levels. Despite no significant changes in overall body weight, microMRI assessments demonstrated a significant increase in visceral and subcutaneous abdominal fat volumes in APE1/Ref-1+/- mice. Conclusion APE1/Ref-1 is crucial in adipokine regulation and mitigating oxidative stress. These findings suggest its involvement in adipose tissue dysfunction, highlighting its potential impact on abdominal fat distribution and its implications for obesity and oxidative stress-related conditions.