Artificial Intelligence Evaluation of the Utility of HALP Score and Hematological Indicators in Estimating No-Reflow After Primary Percutaneous Coronary Intervention in Patients with ST-Segment Elevation Myocardial Infarction

Background: Acute Myocardial Infarction (AMI) is a leading cause of mortality globally, with ST-segment Elevation Myocardial Infarction (STEMI) being a specific type. The study aims to fill the gap in literature regarding the predictive utility of hematological parameters and the HALP score in the context of the "no-reflow" phenomenon in STEMI patients.  To evaluate the predictive efficacy of hematological parameters and the HALP score in identifying the "no-reflow" phenomenon in STEMI patients undergoing Primary Percutaneous Coronary Intervention (PPCI) using Explainable AI (XAI) methodologies. Material and Methods: A retrospective observational design was used, involving 232 STEMI patients who underwent PPCI between January 2020 and September 2023. The cohort was subsequently dichotomized into two subsets based on the presence or absence of the no-reflow phenomenon. Data were collected on demographic variables, MI locations, and hematological parameters. The HALP score was calculated, and XGBoost machine learning models were developed and evaluated. Results: Statistically significant differences were observed in white blood cell count (WBC), monocyte (MO), neutrophil (NEU), platelet (PLT), albumin (ALB), and the MPV/LY ratio (MPVLR) between the 'NORMAL-REFLOW' and 'NO-REFLOW' categories. The XGBoost model showed good performance in the training set but had limitations in sensitivity in the test set. Conclusion: In this study, according to artificial intelligence analysis, the most important hematological parameter in predicting no-relow was MPVLR. However, in this study, HALP score was not found to be effective in predicting no-reflow. The study provides valuable insights into the predictive factors for reflow outcomes in STEMI patients.