Objective: Numerous supervised learning models aimed at classifying 12-lead electrocardiograms into different groups have shown impressive performance by utilizing deep learning algorithms. However, few studies are dedicated to applying the Generative Pre-trained Transformer (GPT) model in interpreting electrocardiogram (ECG) using natural language. Thus, we are pioneering the exploration of this uncharted territory by employing the CardioGPT model to tackle this challenge. Methods: We used a dataset of ECGs (standard 10s, 12-channel format) from adult patients, with 60 distinct rhythms or conduction abnormalities annotated by board-certified, actively practicing cardiologists. The ECGs were collected from The First Affiliated Hospital of Ningbo University and Shanghai East Hospital. The dataset is partitioned into training (80%), validation (10%), and test (10%) cohorts for comprehensive evaluation. Each cohort contains ECGs from distinct patients, considering some patients took repeated ECG measurements. The proposed algorithm is evaluated in two levels, self-performance measurement and comparison with the residual neural network classification model. Two scores are used for self-performance measurement, including Bilingual Evaluation Understudy (BLEU) and Recall-Oriented Understudy for Gisting Evaluation (ROUGE). To compare the performance of the proposed model with the residual neural network model, we assessed the F1 score and area under the receiver operating characteristic curve (AUC). Results: We have observed promising performance metrics across multiple evaluation criteria through an extensive evaluation of a large 12-lead ECG database comprising 1,128,553 ECG readings from 754,920 patients. The CardioGPT model exhibited high BLEU and ROUGE scores with 0.68 (95% CI: 0.66, 0.71) and 0.81 (95% CI: 0.79, 0.84). Furthermore, in the classification performance measurement setting, the CardioGPT achieved an average F1-score of 0.91(95% CI: 0.89, 0.93) and AUC of 0.82(95% CI: 0.79, 0.84) and has higher scores than that of the convolutional neural network model, indicating its proficiency in accurately classifying ECG recordings. Conclusion: By leveraging the power of transformer structure model and natural language processing, the GPT model addresses the challenge of imbalanced learning commonly encountered in ECG classification tasks. The results indicate that the GPT model can accurately interpret ECG using natural language, providing valuable insights into the underlying patterns and abnormalities present in the data. Significance: The pioneering application of the GPT model for interpreting ECGs with natural language demonstrates its potential to address ECG classification challenges and offer valuable insights into cardiac health.

Cardiovascular and other circulatory system diseases have been implicated in the severity of COVID-19 in adults. This study provides a super learner ensemble of models for predicting COVID-19 severity among these patients. The Cerner Real-World Database was used for this study. Data on adult patients (18 years or older) with cardiovascular and related circulatory diseases between 2017 and 2019 were retrieved and a total of 13 these conditions were identified. Among these patients, 33,042 admitted with positive diagnoses for COVID-19 between March 2020 and June 2020 (from 59 hospitals) were identified and selected for this study. A total of 14 statistical and machine learning models were developed and combined into a single more powerful super learning model for predicting COVID-19 severity on admission to the hospital. LASSO regression, a full extreme gradient boosting model with tree depth of 2, and a full logistic regression model were the most predictive with cross-validated AUROCs of 0.7964, 0.7961, and 0.7958 respectively. The resulting super learner ensemble model had a cross validated AUROC of 0.8006 (range: 0.7814, 0.8163). The unbiased AUROC of the super learner model on an independent test set was 0.8057 (95% CI: 0.7954, 0.8159). Highly predictive models can be built to predict COVID-19 severity of patients with cardiovascular and other circulatory conditions. Super learning ensembles will improve individual and classical ensemble models significantly.

ECG (Electrocardiogram) data analysis is one of the most widely used and important tools in cardiology diagnostics. In recent years the development of advanced deep learning techniques and GPU hardware have made it possible to train neural network models that attain exceptionally high levels of accuracy in complex tasks such as heart disease diagnoses and treatments. We investigate the use of ECGs as biometrics in human identification systems by implementing state-of-the-art deep learning models. We train convolutional neural network models on approximately 81k patients from the US, Germany and China. Currently, this is the largest research project on ECG identification. Our models achieved an overall accuracy of 95.69%. Furthermore, we assessed the accuracy of our ECG identification model for distinct groups of patients with particular heart conditions and combinations of such conditions. For example, we observed that the identification accuracy was the highest (99.7%) for patients with both ST changes and supraventricular tachycardia. We also found that the identification rate was the lowest for patients diagnosed with both atrial fibrillation and complete right bundle branch block (49%). We discuss the implications of these findings regarding the reidentification risks of patients based on ECG data and how seemingly anonymized ECG datasets can cause privacy concerns for the patients.

Suicidal and self-injurious incidents in correctional settings deplete the institutional and healthcare resources, create disorder and stress for staff and other inmates. Traditional statistical analyses provide some guidance, but they can only be applied to structured data that are often difficult to collect and their recommendations are often expensive to act upon. This study aims to extract information from medical and mental health progress notes using AI algorithms to make actionable predictions of suicidal and self-injurious events to improve the efficiency of triage for health care services and prevent suicidal and injurious events from happening at California's Orange County Jails. The results showed that the notes data contain more information with respect to suicidal or injurious behaviors than the structured data available in the EHR database at the Orange County Jails. Using the notes data alone (under-sampled to 50%) in a Transformer Encoder model produced an AUC-ROC of 0.862, a Sensitivity of 0.816, and a Specificity of 0.738. Incorporating the information extracted from the notes data into traditional Machine Learning models as a feature alongside structured data (under-sampled to 50%) yielded better performance in terms of Sensitivity (AUC-ROC: 0.77, Sensitivity: 0.89, Specificity: 0.65). In addition, under-sampling is an effective approach to mitigating the impact of the extremely imbalanced classes.

Background: Although prior research has estimated the overarching cost burden of heart failure (HF), a thorough analysis examining medical expense differences and trends, specifically among commercially insured patients with heart failure, is still lacking. Thus, the study aims to examine historical trends and differences in medical costs for commercially insured heart failure patients in the United States from 2006 to 2021. Methods: A population-based, cross-sectional analysis of medical and pharmacy claims data (IQVIA PharMetrics^® Plus for Academic) from 2006 to 2021 was conducted. The cohort included adult patients (age > = 18) who were enrolled in commercial insurance plans and had healthcare encounters with a primary diagnosis of HF. The primary outcome measures were the average total annual payment per patient and per cost categories encompassing hospitalization, surgery, emergency department (ED) visits, outpatient care, post-discharge care, and medications. The sub-group measures included systolic, diastolic, and systolic combined with diastolic, age, gender, comorbidity, regions, states, insurance payment, and self-payment. Results: The study included 422,289 commercially insured heart failure (HF) patients in the U.S. evaluated from 2006 to 2021. The average total annual cost per patient decreased overall from $9,636.99 to $8,201.89, with an average annual percentage change (AAPC) of -1.11% (95% CI: -2% to -0.26%). Hospitalization and medication costs decreased with an AAPC of -1.99% (95% CI: -3.25% to -0.8%) and − 3.1% (95% CI: -6.86–0.69%). On the other hand, post-discharge, outpatient, ED visit, and surgery costs increased by an AAPC of 0.84% (95% CI: 0.12–1.49%), 4.31% (95% CI: 1.03–7.63%), 7.21% (95% CI: 6.44–8.12%), and 9.36% (95% CI: 8.61–10.19%). Conclusions: The study’s findings reveal a rising trend in average total annual payments per patient from 2006 to 2015, followed by a subsequent decrease from 2016 to 2021. This decrease was attributed to the decline in average patient costs within the Medicare Cost insurance category after 2016, coinciding with the implementation of the Medicare Access and CHIP Reauthorization Act (MACRA) of 2015. Additionally, expenses related to surgical procedures, emergency department (ED) visits, and outpatient care have shown substantial growth over time. Moreover, significant differences across various variables have been identified.