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: Intellectual disability (ID) may be associated with an increased risk of diabetes mellitus (DM). However, evidence from longitudinal studies is scarce, particularly in Asian populations. Methods: This retrospective cohort study used representative linked data from the Korea National Disability Registration System and the National Health Insurance Service database. Adults (≥20 years) who received a national health examination in 2009 (3,385 individuals with ID and 3,463,604 individuals without ID) were included and followed until 2020. ID was identified using legal registration information. Incident DM was defined by prescription records with relevant diagnostic codes. Multivariable-adjusted Cox proportional hazards regression models were used to estimate the adjusted hazard ratio (aHR) and 95% confidence interval (CI) for DM risks in individuals with ID compared to those without ID. Results: Over a mean follow-up of 9.8 years, incident DM occurred in 302 (8.9%) individuals with ID and 299,156 (8.4%) individuals without ID. Having ID was associated with increased DM risk (aHR, 1.38; 95% CI, 1.23 to 1.55). Sensitivity analysis confirmed a higher DM risk in individuals with ID (aHR, 1.39; 95% CI, 1.24 to 1.56) than those with other disabilities (aHR, 1.11; 95% CI, 1.10 to 1.13) or no disability (reference). Stratified analysis showed higher DM risk in non-hypertensive subjects (aHR, 1.63; 95% CI, 1.43 to 1.86) compared to hypertensive subjects (aHR, 1.00; 95% CI, 0.80 to 1.26; P for interaction <0.001). Conclusion Adults with ID have an increased risk of developing DM, highlighting the need for targeted public health strategies to promote DM prevention in this population.

Purpose: Lupus nephritis (LN) can be caused by the complement activation. This study aimed to investigate the differences and clinical implications of the activation pattern of the complement system for pediatric and adult LN patients. Methods: We retrospectively reviewed the medical records of 40 patients (14 pediatric and 26 adult patients) diagnosed with LN through kidney biopsy. Results: The mean ages at diagnosis of pediatric and adult patients were 11.7±2.92 and 37.3±13.5 years, respectively. At the first LN diagnosis, compared with adult patients, pediatric patients had a higher estimated glomerular filtration rate and milder proteinuria; however, there was no statistical significance. The age-adjusted mean serum complement 3 value was significantly lower in the pediatric group (33.0±11.3 mg/dL) than in the adult group (50.8±25.2 mg/dL) (P<0.01). Based on the findings of kidney biopsy, no significant differences were observed in the severity of pathologic classification and the positive rate of complements between adults and children. However, the chronicity index score of adult patients was significantly higher than that of pediatric patients and in the case of complement 4d, despite a similar positive rate, the intensity was significantly stronger for adults (2.35±0.83 vs. 1.54±0.52, (P=0.04). Conclusions The activation pattern of the complement system in LN differs clinicopathologically between pediatric and adult patients and these differences might play an important role in the age-dependent prognosis of LN.

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: Intellectual disability (ID) may be associated with an increased risk of diabetes mellitus (DM). However, evidence from longitudinal studies is scarce, particularly in Asian populations. Methods: This retrospective cohort study used representative linked data from the Korea National Disability Registration System and the National Health Insurance Service database. Adults (≥20 years) who received a national health examination in 2009 (3,385 individuals with ID and 3,463,604 individuals without ID) were included and followed until 2020. ID was identified using legal registration information. Incident DM was defined by prescription records with relevant diagnostic codes. Multivariable-adjusted Cox proportional hazards regression models were used to estimate the adjusted hazard ratio (aHR) and 95% confidence interval (CI) for DM risks in individuals with ID compared to those without ID. Results: Over a mean follow-up of 9.8 years, incident DM occurred in 302 (8.9%) individuals with ID and 299,156 (8.4%) individuals without ID. Having ID was associated with increased DM risk (aHR, 1.38; 95% CI, 1.23 to 1.55). Sensitivity analysis confirmed a higher DM risk in individuals with ID (aHR, 1.39; 95% CI, 1.24 to 1.56) than those with other disabilities (aHR, 1.11; 95% CI, 1.10 to 1.13) or no disability (reference). Stratified analysis showed higher DM risk in non-hypertensive subjects (aHR, 1.63; 95% CI, 1.43 to 1.86) compared to hypertensive subjects (aHR, 1.00; 95% CI, 0.80 to 1.26; P for interaction <0.001). Conclusion Adults with ID have an increased risk of developing DM, highlighting the need for targeted public health strategies to promote DM prevention in this population.

Purpose: Lupus nephritis (LN) can be caused by the complement activation. This study aimed to investigate the differences and clinical implications of the activation pattern of the complement system for pediatric and adult LN patients. Methods: We retrospectively reviewed the medical records of 40 patients (14 pediatric and 26 adult patients) diagnosed with LN through kidney biopsy. Results: The mean ages at diagnosis of pediatric and adult patients were 11.7±2.92 and 37.3±13.5 years, respectively. At the first LN diagnosis, compared with adult patients, pediatric patients had a higher estimated glomerular filtration rate and milder proteinuria; however, there was no statistical significance. The age-adjusted mean serum complement 3 value was significantly lower in the pediatric group (33.0±11.3 mg/dL) than in the adult group (50.8±25.2 mg/dL) (P<0.01). Based on the findings of kidney biopsy, no significant differences were observed in the severity of pathologic classification and the positive rate of complements between adults and children. However, the chronicity index score of adult patients was significantly higher than that of pediatric patients and in the case of complement 4d, despite a similar positive rate, the intensity was significantly stronger for adults (2.35±0.83 vs. 1.54±0.52, (P=0.04). Conclusions The activation pattern of the complement system in LN differs clinicopathologically between pediatric and adult patients and these differences might play an important role in the age-dependent prognosis of LN.

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: Intellectual disability (ID) may be associated with an increased risk of diabetes mellitus (DM). However, evidence from longitudinal studies is scarce, particularly in Asian populations. Methods: This retrospective cohort study used representative linked data from the Korea National Disability Registration System and the National Health Insurance Service database. Adults (≥20 years) who received a national health examination in 2009 (3,385 individuals with ID and 3,463,604 individuals without ID) were included and followed until 2020. ID was identified using legal registration information. Incident DM was defined by prescription records with relevant diagnostic codes. Multivariable-adjusted Cox proportional hazards regression models were used to estimate the adjusted hazard ratio (aHR) and 95% confidence interval (CI) for DM risks in individuals with ID compared to those without ID. Results: Over a mean follow-up of 9.8 years, incident DM occurred in 302 (8.9%) individuals with ID and 299,156 (8.4%) individuals without ID. Having ID was associated with increased DM risk (aHR, 1.38; 95% CI, 1.23 to 1.55). Sensitivity analysis confirmed a higher DM risk in individuals with ID (aHR, 1.39; 95% CI, 1.24 to 1.56) than those with other disabilities (aHR, 1.11; 95% CI, 1.10 to 1.13) or no disability (reference). Stratified analysis showed higher DM risk in non-hypertensive subjects (aHR, 1.63; 95% CI, 1.43 to 1.86) compared to hypertensive subjects (aHR, 1.00; 95% CI, 0.80 to 1.26; P for interaction <0.001). Conclusion Adults with ID have an increased risk of developing DM, highlighting the need for targeted public health strategies to promote DM prevention in this population.

Purpose: Lupus nephritis (LN) can be caused by the complement activation. This study aimed to investigate the differences and clinical implications of the activation pattern of the complement system for pediatric and adult LN patients. Methods: We retrospectively reviewed the medical records of 40 patients (14 pediatric and 26 adult patients) diagnosed with LN through kidney biopsy. Results: The mean ages at diagnosis of pediatric and adult patients were 11.7±2.92 and 37.3±13.5 years, respectively. At the first LN diagnosis, compared with adult patients, pediatric patients had a higher estimated glomerular filtration rate and milder proteinuria; however, there was no statistical significance. The age-adjusted mean serum complement 3 value was significantly lower in the pediatric group (33.0±11.3 mg/dL) than in the adult group (50.8±25.2 mg/dL) (P<0.01). Based on the findings of kidney biopsy, no significant differences were observed in the severity of pathologic classification and the positive rate of complements between adults and children. However, the chronicity index score of adult patients was significantly higher than that of pediatric patients and in the case of complement 4d, despite a similar positive rate, the intensity was significantly stronger for adults (2.35±0.83 vs. 1.54±0.52, (P=0.04). Conclusions The activation pattern of the complement system in LN differs clinicopathologically between pediatric and adult patients and these differences might play an important role in the age-dependent prognosis of LN.

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.