Table 1 Variables included in construction of anesthesiology decision analysis platform
From: Anesthesia decision analysis using a cloud-based big data platform
Feature type | Features | Preprocessing |
|---|---|---|
Patient characteristics | ||
Continuous | Age, height, weight, BSA, BMI and Tumor size | Normalization |
Categorical | Sex, Charlson Comorbidity Index, Mallampati grading, mouth opening, Reoperation, tumor stage and differentiation, Metastasis, functional capacity, ASA physical status, ASA, emergency status, anesthesia type, and surgery type | One-hot encoding |
Categorical comorbid conditions | Hypertension, coronary artery disease, prior myocardial infarction, congestive heart failure, diastolic function, left ventricular ejection fraction, aortic stenosis, atrial fibrillation, prior stroke or transient ischemic attack, pacemaker or implanted defibrillator, peripheral artery disease, deep venous thrombosis, pulmonary embolism, diabetes, chronic kidney disease, ongoing dialysis, pulmonary hypertension, chronic obstructive pulmonary disease, asthma, obstructive sleep apnea, cirrhosis, any cancer, gastroesophageal reflux, anemia, positive Coombs test result, dementia, ever-smoker, and alcohol addiction | One-hot encoding if not binary |