Background/Aims: Atrial fibrillation (AF) significantly contributes to global morbidity and mortality. Paroxysmal atrial fibrillation (PAF) is particularly common among patients with cryptogenic strokes or transient ischemic attacks and has a silent nature. This study aims to develop reliable artificial intelligence (AI) algorithms to detect early signs of AF in patients with normal sinus rhythm (NSR) using a 12-lead electrocardiogram (ECG). Methods: Between 2013 and 2020, 552,372 ECG traces from 318,321 patients were collected and split into training (n = 331,422), validation (n = 110,475), and test sets (n = 110,475). Deep neural networks were then trained to predict AF onset within one month of NSR. Model performance was evaluated using the area under the receiver operating characteristic curve (AUROC). An explainable AI technique was employed to identify the inference evidence underlying the predictions of deep learning models. Results: The AUROC for early diagnosis of PAF was 0.905 ± 0.007. The findings reveal that the vicinity of the T wave, including the ST segment and S-peak, significantly influences the ability of the trained neural network to diagnose PAF. Additionally, comparing the summarized ECG in NSR with those in PAF revealed that nonspecific ST-T abnormalities and inverted T waves were associated with PAF. Conclusions Deep learning can predict AF onset from NSR while detecting key features that influence decisions. This suggests that identifying undetected AF may serve as a predictive tool for PAF screening, offering valuable insights into cardiac dysfunction and stroke risk.

Background/Aims: Atrial fibrillation (AF) significantly contributes to global morbidity and mortality. Paroxysmal atrial fibrillation (PAF) is particularly common among patients with cryptogenic strokes or transient ischemic attacks and has a silent nature. This study aims to develop reliable artificial intelligence (AI) algorithms to detect early signs of AF in patients with normal sinus rhythm (NSR) using a 12-lead electrocardiogram (ECG). Methods: Between 2013 and 2020, 552,372 ECG traces from 318,321 patients were collected and split into training (n = 331,422), validation (n = 110,475), and test sets (n = 110,475). Deep neural networks were then trained to predict AF onset within one month of NSR. Model performance was evaluated using the area under the receiver operating characteristic curve (AUROC). An explainable AI technique was employed to identify the inference evidence underlying the predictions of deep learning models. Results: The AUROC for early diagnosis of PAF was 0.905 ± 0.007. The findings reveal that the vicinity of the T wave, including the ST segment and S-peak, significantly influences the ability of the trained neural network to diagnose PAF. Additionally, comparing the summarized ECG in NSR with those in PAF revealed that nonspecific ST-T abnormalities and inverted T waves were associated with PAF. Conclusions Deep learning can predict AF onset from NSR while detecting key features that influence decisions. This suggests that identifying undetected AF may serve as a predictive tool for PAF screening, offering valuable insights into cardiac dysfunction and stroke risk.

Background/Aims: Atrial fibrillation (AF) significantly contributes to global morbidity and mortality. Paroxysmal atrial fibrillation (PAF) is particularly common among patients with cryptogenic strokes or transient ischemic attacks and has a silent nature. This study aims to develop reliable artificial intelligence (AI) algorithms to detect early signs of AF in patients with normal sinus rhythm (NSR) using a 12-lead electrocardiogram (ECG). Methods: Between 2013 and 2020, 552,372 ECG traces from 318,321 patients were collected and split into training (n = 331,422), validation (n = 110,475), and test sets (n = 110,475). Deep neural networks were then trained to predict AF onset within one month of NSR. Model performance was evaluated using the area under the receiver operating characteristic curve (AUROC). An explainable AI technique was employed to identify the inference evidence underlying the predictions of deep learning models. Results: The AUROC for early diagnosis of PAF was 0.905 ± 0.007. The findings reveal that the vicinity of the T wave, including the ST segment and S-peak, significantly influences the ability of the trained neural network to diagnose PAF. Additionally, comparing the summarized ECG in NSR with those in PAF revealed that nonspecific ST-T abnormalities and inverted T waves were associated with PAF. Conclusions Deep learning can predict AF onset from NSR while detecting key features that influence decisions. This suggests that identifying undetected AF may serve as a predictive tool for PAF screening, offering valuable insights into cardiac dysfunction and stroke risk.

Background/Aims: Atrial fibrillation (AF) significantly contributes to global morbidity and mortality. Paroxysmal atrial fibrillation (PAF) is particularly common among patients with cryptogenic strokes or transient ischemic attacks and has a silent nature. This study aims to develop reliable artificial intelligence (AI) algorithms to detect early signs of AF in patients with normal sinus rhythm (NSR) using a 12-lead electrocardiogram (ECG). Methods: Between 2013 and 2020, 552,372 ECG traces from 318,321 patients were collected and split into training (n = 331,422), validation (n = 110,475), and test sets (n = 110,475). Deep neural networks were then trained to predict AF onset within one month of NSR. Model performance was evaluated using the area under the receiver operating characteristic curve (AUROC). An explainable AI technique was employed to identify the inference evidence underlying the predictions of deep learning models. Results: The AUROC for early diagnosis of PAF was 0.905 ± 0.007. The findings reveal that the vicinity of the T wave, including the ST segment and S-peak, significantly influences the ability of the trained neural network to diagnose PAF. Additionally, comparing the summarized ECG in NSR with those in PAF revealed that nonspecific ST-T abnormalities and inverted T waves were associated with PAF. Conclusions Deep learning can predict AF onset from NSR while detecting key features that influence decisions. This suggests that identifying undetected AF may serve as a predictive tool for PAF screening, offering valuable insights into cardiac dysfunction and stroke risk.

Background/Aims: Atrial fibrillation (AF) significantly contributes to global morbidity and mortality. Paroxysmal atrial fibrillation (PAF) is particularly common among patients with cryptogenic strokes or transient ischemic attacks and has a silent nature. This study aims to develop reliable artificial intelligence (AI) algorithms to detect early signs of AF in patients with normal sinus rhythm (NSR) using a 12-lead electrocardiogram (ECG). Methods: Between 2013 and 2020, 552,372 ECG traces from 318,321 patients were collected and split into training (n = 331,422), validation (n = 110,475), and test sets (n = 110,475). Deep neural networks were then trained to predict AF onset within one month of NSR. Model performance was evaluated using the area under the receiver operating characteristic curve (AUROC). An explainable AI technique was employed to identify the inference evidence underlying the predictions of deep learning models. Results: The AUROC for early diagnosis of PAF was 0.905 ± 0.007. The findings reveal that the vicinity of the T wave, including the ST segment and S-peak, significantly influences the ability of the trained neural network to diagnose PAF. Additionally, comparing the summarized ECG in NSR with those in PAF revealed that nonspecific ST-T abnormalities and inverted T waves were associated with PAF. Conclusions Deep learning can predict AF onset from NSR while detecting key features that influence decisions. This suggests that identifying undetected AF may serve as a predictive tool for PAF screening, offering valuable insights into cardiac dysfunction and stroke risk.

Proximal junctional problems are among the potential complications of surgery for adult spinal deformity (ASD) and are associated with higher morbidity and increased rates of revision surgery. The diverse manifestations of proximal junctional problems range from proximal junctional kyphosis (PJK) to proximal junctional failure (PJF). Although there is no universally accepted definition for PJK, the most common is a proximal junctional angle greater than 10° that is at least 10° greater than the preoperative measurement. PJF represents a progression from PJK and is characterized by pain, gait disturbances, and neurological deficits. The risk factors for PJK can be classified according to patient-related, radiological, and surgical factors. Based on an understanding of the modifiable factors that contribute to reducing the risk of PJK, prevention strategies are critical for patients with ASD.

We find out the clusters with high toxoplasmosis risk to discuss the geographical pattern in Gyodong-myeon and Samsan-myeon of Ganghwa-gun, Cheorwon-gun, and Goseong-gun, Korea. Seroepidemiological data of toxoplasmosis surveyed using rapid diagnostic tests for the residents in the areas in 2019 were analyzed to detect clusters of the infection. The cluster was investigated using the SaTScan program which is based on Kulldorff’s scan statistic. The clusters were found with P-values in each region analyzed in the program, and the risk and patient incidence of specific areas can be examined by the values such as relative risk and log likelihood ratio. Jiseok-ri and Insa-ri were found to be a cluster in Gyodong-myeon and Seokmo-ri was the cluster in Samsan-myeon. Yangji-ri and Igil-ri were found to be a cluster in Cheorwon-gun and Madal-ri and Baebong-ri were the cluster in Goseong-gun. This findings can be used to monitor and prevent toxoplasmosis infections occurring in vulnerable areas.

We find out the clusters with high toxoplasmosis risk to discuss the geographical pattern in Gyodong-myeon and Samsan-myeon of Ganghwa-gun, Cheorwon-gun, and Goseong-gun, Korea. Seroepidemiological data of toxoplasmosis surveyed using rapid diagnostic tests for the residents in the areas in 2019 were analyzed to detect clusters of the infection. The cluster was investigated using the SaTScan program which is based on Kulldorff’s scan statistic. The clusters were found with P-values in each region analyzed in the program, and the risk and patient incidence of specific areas can be examined by the values such as relative risk and log likelihood ratio. Jiseok-ri and Insa-ri were found to be a cluster in Gyodong-myeon and Seokmo-ri was the cluster in Samsan-myeon. Yangji-ri and Igil-ri were found to be a cluster in Cheorwon-gun and Madal-ri and Baebong-ri were the cluster in Goseong-gun. This findings can be used to monitor and prevent toxoplasmosis infections occurring in vulnerable areas.

Background: Firefighters inevitably encounter emotionally and physically stressful situations at work. Even firefighters without diagnosed post-traumatic stress disorder receive clinical attention because the nature of the profession exposes them to repetitive trauma and high occupational stress. This study investigated gray matter abnormalities related to high occupational stress in firefighters using voxel-based morphometry (VBM) and surface-based morphometry (SBM). Methods: We assessed 115 subjects (112 males and 3 females) using magnetic resonance imaging and evaluated occupational stress by the Korean Occupational Stress Scale-26 (KOSS-26). Subjects were classified into highly or lowly stressed groups based on the median value of the KOSS-26. Results: In VBM analysis, we found that firefighters with high occupational stress had lower gray matter volume (GMV) in both sides of the insula, the left amygdala, the right medial prefrontal cortex (mPFC), and the anterior cingulate cortex than firefighters with low occupational stress. In SBM analysis based on regions of interest, the GMV of the bilateral insula and right mPFC were also lower in the highly stressed group. Within the highly stressed group, low GMV of the insula was significantly correlated with the length of service (left: r = −0.347, P = 0.009; right: r = −0.333, P = 0.012). Conclusion Our findings suggest that regional GMV abnormalities are related to occupational stress. Regional gray matter abnormalities and related emotional dysregulation may contribute to firefighter susceptibility to burnout.

The purpose of this study is to find out the clusters with high toxoplasmosis risk to discuss the geographical pattern in 2 islands of Gyodong-myeon and Samsan-myeon in Ganghwa-gun, Korea. Seroepidemiological data of toxoplasmosis surveyed using rapid diagnostic tests for the residents in 2 islands from 2010 to 2013 were analyzed to detect clusters of the infection. The cluster was investigated using the SatScan program which is based on Kulldorff’s scan statistic. The clusters were found with P-values in each region analyzed in the program, and the risk and patient incidence of specific areas can be examined by the values such as relative risk and log likelyhood ratio. Jiseok-ri was found to be a cluster in Gyodong-myeon and Ha-ri was the cluster in Samsan-myeon. This findings can be used to monitor and prevent toxoplasmosis infections occurring in vulnerable areas.