Polskie Archiwum Medycyny Wewnętrznej
-
Pol. Arch. Med. Wewn. · Mar 2024
Practical use case of natural language processing for observational clinical research data retrieval from electronic health records: AssistMED project.
Electronic health records (EHR) contain data valuable for clinical research but in textual format, requiring encoding to databases by a human- a lengthy and costly process. Natural language processing (NLP) is a computational technique that allows text analysis. ⋯ NLP utilization on EHR may accelerate acquisition and provide accurate data for a retrospective study.
-
Pol. Arch. Med. Wewn. · Mar 2024
A comparison of interpretable XGBoost and artificial neural network model for the prediction of severe acute pancreatitis.
Acute pancreatitis (AP) that progresses to persistent organ failure is defined as severe acute pancreatitis (SAP) which has a relatively high mortality. Early establishment of a prediction model is crucial for the improvement of disease prognosis. ⋯ An interpretable XGBoost model showed higher discriminatory efficiency in predicting SAP compared to ANN.
-
Pol. Arch. Med. Wewn. · Mar 2024
Unattended automatic blood pressure measurements versus conventional office readings in predicting hypertension-mediated organ damage.
Hypertension is a leading cardiovascular risk factor. Accurate blood pressure measurement is pivotal in hypertension diagnosis and management. Conventional office measurements (OBP) are error-prone, exacerbated by the white coat effect. Unattended automated office blood pressure measurement (UAOBP) is emerging as an alternative, mitigating the white coat effect. However, its ability to predict hypertension-mediated organ damage (HMOD) remains disputable. ⋯ The UAOBP did not prove superior in predicting HMOD compared to OBP, Further research is warranted to determine the role of UAOBP in clinical practice.