Imputation of missing data by R MICE

I have completed the process in terms of reaching the end of the steps I've found on numerous tutorial sites/videos however I'm not quite satisfied - how does 10 different imputation datasets and a pooled step-wise regression model give me the final dataset? None of the tutorials address this question. Perhaps its obvious to you guys.


Less is more. Stay pure. Stay poor.
Typically you just run your target model on each imputed set and pool estimates. I don't know about the stepwise model?