Use of classifiers to optimise the identification and characterisation of metastatic breast cancer in a nationwide administrative registry.

To our knowledge, this is the first study to use machine learning to identify MBC regardless of stage at diagnosis in health registries covering the entire population of Sweden.