When Scott Isbell, PhD, pitched his talk show format for the 2020 AACC Annual Scientific Meeting, he had no idea that he wouldn’t be able to run down the aisles mic-in-hand, talking artificial intelligence and cardiovascular lab medicine (think Neil deGrasse Tyson meets Phil Donahue).

However, as we’ve learned in the last few months watching our favorite late-night hosts continue to interview A-listers, talk shows translate very well to a virtual format. Expect no less from Isbell and the A(ACC)-listers he’s lined up for the special session on Sunday, December 13 at 3:00 pm Central: “Today in Clinical Chemistry – AACCs Annual Scientific Meeting Talk Show on Machine Learning and MI”.

This show is a can’t-miss session for AACC members interested in real world applications for machine learning in lab medicine. Isbell hosts an interactive session asking probing questions and engaging with our live (virtual) audience. The specific real-world application is myocardial infarction; the guest experts are none other than Fred Apple, PhD, renowned for his work on troponin and other cardiac biomarkers, and interventional cardiologist Yader Sandoval, MD.

Sandoval is a physician-investigator with a research program aimed at improving the diagnosis, management, and outcomes of patients that present with acute coronary syndromes or are burdened with coronary artery disease. He is particularly interested in improving the diagnosis of acute myocardial infarction, and has helped develop and validate novel diagnostic strategies using high-sensitivity cardiac troponin to improve the rapid rule-in and rule-out of acute myocardial infarction.

Apple is a key researcher at Hennepin Healthcare Cardiac Biomarker Trials Laboratory in Minneapolis, Minnesota. Apple and his lab have been involved in hundreds of clinical trials and analytical evaluations focused on biomarkers of myocardial injury. Apple is internationally recognized as an expert in cardiac biomarkers, and his laboratory serves as a center of excellence for forensic and clinical toxicology applied research studies.

Using a training dataset of age, sex, and troponin values from more than 3,000 patients, Apple and Sandoval validated a machine learning algorithm in a cohort of nearly 8,000 patients. The results were recently published in Circulation and reported that the algorithm correctly identified patients suffering an acute myocardial infarction with superior performance to conventional diagnostic strategies.

“Machine learning is all the rage,” Isbell says. “After the show attendees will be able to articulate the benefits and challenges of applying machine learning techniques in the lab using the example of myocardial infarction as a case study.”