Elana Gotkin
Thursday, June 1, 2023 (HealthDay News) — Machine learning models incorporating cardiac troponin levels and clinical features may improve diagnosis of myocardial infarction, announced online May 11. research results have been published. natural medicine.
Dimitrios Doudesis, Ph.D., University of Edinburgh, UK, et al. developed a machine learning model that integrates cardiac troponin concentrations and clinical features at presentation or serial examination to compute a collaboration for the diagnosis and assessment of acute coronary arteries. Syndrome to identify possible myocardial infarction (CoDE-ACS) score. The model was trained on data from 10,038 He patients and validated externally using data from 10,286 He patients in seven cohorts.
The researchers found that CoDE-ACS had excellent discrimination of myocardial infarction (area under the curve, 0.953) and performed well across subgroups. Compared with fixed myocardial troponin thresholds, CoDE-ACS identified more patients as unlikely to have myocardial infarction at presentation (61 vs. 27 percent) with similar negative predictive values Fewer patients identified as likely (10 vs. 16 percent), yielding a higher positive predictive value. Patients identified as having a low risk of myocardial infarction compared with those with an intermediate or high probability had cardiac mortality was low. .
Others are reading…
“If adopted, CoDE-ACS could reduce time spent in the emergency room, prevent unnecessary hospitalizations, and improve early treatment of myocardial infarction, benefiting both patients and providers. can lead to,” the authors write.
Several authors clarified ties to the biopharmaceutical industry.
