An AI model to assess the risk-reward ratio of heart surgery

Researchers from premier institutions around the world have built an AI predictive model to help doctors guide heart patients better and help decide on the need for surgery on a case-to-case basis. Patients with existing comorbidities are at a higher risk for complications such as stroke, prolonged ventilation, renal failure and death after mitral valve replacement surgery. 

The study titled, “Machine learning models for mitral valve replacement: A comparative analysis with the Society of Thoracic Surgeons risk score”,  proposed risk models that complement existing STS (Society of Thoracic Surgeons) models in predicting mortality, prolonged ventilation and renal failure, allowing healthcare providers to more accurately assess a patient’s risk of morbidity and mortality when undergoing mitral valve surgery.

The existing STS risk models for predicting outcomes of mitral valve surgery assume a linear and cumulative impact of variables.

The data–from the STS Adult Cardiac Surgery Database for MVS from 2008 to 2017– included 383,550 procedures and 89 variables. ML algorithms were used to train models to predict postoperative outcomes for MVS patients. Each model's discrimination and calibration performance were validated using unseen data against the STS risk score.

Comprehensive mortality and morbidity risk assessment scores were derived from a training set of 287,662 observations. The area under the curve (AUC) for mortality ranged from 0.77 to 0.83, leading to a 3% increase in predictive accuracy compared to the STS score. The ML models led to improved calibration performance for mortality, prolonged ventilation, and renal failure, especially in cases of reconstruction/repair and replacement surgery.