Evaluating Artificial Intelligence Algorithms Using Blood Cell Data to Predict Acute Coronary Syndrome in Adults An Observational Study Comparing AI and Cardiac Troponin Testing

What this Trial Is About

Researchers are evaluating whether artificial intelligence AI algorithms can predict or exclude acute coronary syndrome ACS in adults by analyzing white blood cell WBC data from routine blood tests. The study also explores whether data on red blood cells erythrocytes and platelets thrombocytes can enhance this prediction. The diagnostic accuracy of AI models will be compared with high-sensitivity cardiac troponin hs-cTn testing, and combinations of AI and hs-cTn will also be assessed. The study is observational and involves two phases training and testing AI models. During training, participants are grouped based on their clinical diagnosis and angiography results, including control, case, supplementary, and rule-out cohorts. WBC data, along with erythrocyte and thrombocyte data when available, will be used to build and refine the AI models. Testing uses a separate all-comer cohort of emergency department patients with suspected ACS to evaluate AI performance without overlap with training data. Participants undergo routine clinical assessments, blood tests, ECGs, and, if indicated, coronary angiography as part of their standard care. The study uses existing clinical and laboratory data to develop AI models, with no additional treatments or procedures required. Researchers will analyze diagnostic accuracy using statistical methods like ROC curves and compare AI results with hs-cTn measurements. Data quality and integrity are ensured through verification and management procedures throughout the study, which is expected to continue for 36 months for both training and testing phases.

Diagnostic Performance of Artificial Intelligence Algorithms in Prediction of Acute Coronary Syndrome Based on White Blood Cell Properties