Neural Network Model as the Team Member in Decision Support

Buzaev, I. V., Plechev, V. V., Nikolaeva, I. E., & Galimova, R. (2017). TCTAP A-048 Neural Network Model as the Multidisciplinary Team Member in Clinical Decision Support to Avoid Medical Mistakes (aLYNX concept) // Journal of the American College of Cardiology, 69(16 Supplement), S25-S26.»

¹Republican Centre of Cardiovascular Diseases, Bashkir State Medical University, Russian Federation

BACKGROUND The feedback is the essential part of any system. aLYNX concept is an idea to use some fuzzy logic algorithm, for example, a neural network model in decision-making system to avoid possible mistakes in a choice between PCI and CABG.

METHODS aLYNX system contains:

first — a registry with parameters, decisions, and late results;

second — machine learning process based on successful cases registry data;

third — the use of the machine learning results as the adviser.

Objective: To show a possibility to build a mathematic model as an adviser for making a decision between CABG and PCI based on the experience of 5107 patients.

RESULTS The neural network was trained by 4,679 patients who achieved 5-year survival. Among them, 2,390 patients underwent

PCI and 2289 CABG. After training, the correlation coefficient (r) of the network was 0.74 for training, 0.67 for validation, 0.71 for test and

0.73 for a total. Simulation of the neural network function has been performed after training in the two groups of patients with a known 5- year outcome. The disagreement rate was significantly higher in the dead patient group than that in the survivor group between a neural network model and heart team [16.8% (787/4679) vs. 20.3% (87/428).