Novel Generative Recurrent Neural Network Framework to Produce Accurate, Applicable, and Deidentified Synthetic Medical Data for Patients With Metastatic Cancer.

PURPOSE: Sensitive patient data cannot be easily shared/analyzed, severely limiting the innovative progress of research, specifically for marginalized/under-represented populations. Existing methods of deidentification are subject to data breaches. The objective of this study was to develop a neural network capable of generating a synthetic version of data for patients with novel postoperative metastatic cancer. METHODS: We analyzed a metastatic cancer patient cohort of 167,474 patients obtained from the National Surgical Quality Improvement Program. Twenty-seven clinical features were analyzed. We created a volume-matched synthetic cohort of 167,474 patients and a reduced-size synthetic cohort of 5,000 patients. The volume-matched and reduced-size synthetic cohorts were compared against the ground truth data to analyze differences in principal component distribution, underlying statistical properties/associations, intervariable correlations, and machine learning classifier performance when developed on the synthetic data. RESULTS: Among 167,474 patients with metastatic cancer in the original data, 50,669 (30.3%) died within 30 days of their index surgery. Our model was able to accurately capture underlying statistical properties, principal components, and intervariable correlations within the ground truth data, yielding an accuracy of 93.2% with a loss of 0.21%, and develop synthetic data capable of training accurate machine learning classifiers. The reduced-size synthetic data accurately replicated all categorical variables and every continuous variable with statistically similar records (P > .05), with the sole exception of preoperative albumin (P < .05). The volume-matched synthetic data frame was able to accurately replicate all categorical variables (P > .05). CONCLUSION: This described methodology can be applied to any structured medical data from any setting, significantly expedite scientific analysis/innovation, and be used to develop improved predictive classifiers with boosted tree-based algorithms, serving as the potential new gold standard of medical data sharing and data augmentation.

Machine learning models to prognose 30-Day Mortality in Postoperative Disseminated Cancer Patients.

Patients with disseminated cancer at higher risk for postoperative mortality see improved outcomes with altered clinical management. Being able to risk stratify patients immediately after their index surgery to flag high risk patients for healthcare providers is vital. The combination of physician uncertainty and a demonstrated optimism bias often lead to an overestimation of patient life expectancy which can precent proper end of life counseling and lead to inadequate postoperative follow up. In this cohort study of 167,474 postoperative patients with multiple types of disseminated cancer, patients at high risk of 30-day postoperative mortality were accurately identified using our machine learning models based solely on clinical features and preoperative lab values. Extreme Gradient Boosting, Random Forest, and Logistic Regression machine learning models were developed on the cohort. Among 167,474 disseminated cancer patients, 50,669 (30.3%) died within 30 days of their index surgery; After preprocessing, 28 features were included in the model development. The cohort was randomly divided into 133,979 patients (80%) for training the models and 33,495 patients (20%) for testing. The extreme gradient boosting model had an AUC of 0.93 (95% CI: 0.926-0.931), the random forest model had an AUC of 0.93 (95% CI: 0.930-0.934), and the logistic regression model had an AUC of 0.90 (95% CI: 0.900-0.906 the index operation. Ultimately, Machine learning models were able to accurately predict short-term postoperative mortality among a heterogenous population of disseminated cancer patients using commonly accessible medical features. These models can be included in electronic health systems to guide clinical judgements that affect direct patient care, particularly in low-resource settings.