Comprehensive Summary
During laparoscopic surgery, there exists a risk of the patient undergoing hypothermia which leads to a number of dangerous health effects during surgery. This paper by Su et al. looks to use machine learning to help predict the risk of a patient experiencing hypothermia to increase preparedness for the surgery. The researchers created several machine learning models and identified the best performing to help physicians make decisions that impact patient care. Their study looked at 1,030 laparoscopy surgery patient cases and used the Least Absolute Shrinkage and Selection Operator (LASSO) algorithm to identify the most important factors that impact hypothermia risk. Their findings highlighted the main importance of 4 factors: operating room temperature, rinsing fluids during the operation, volume of CO2, and blood loss. From these factors, they trained several machine learning models on the patient case data. Finally, they performed analysis on the results of all the models and settled on the best model with an accuracy of 76.2%.
Outcomes and Implications
This topic has been studied using machine learning in similar ways in the past. The researchers highlight the existing need to perform more interpretive studies using such models as that would be much more helpful and clinically relevant. This paper advances into this territory by creating a more comprehensive predictive model which considers many factors and how they exist together. The researchers argue that the furtherance of this work can help clinicians make tailored decisions to change factors which could specifically reduce the risk for hypothermia during surgery. Further studies are needed to apply this research directly to the operating room, but Su et al. presents another step closer to clinical implementation.