Within a typical hospital information system many different subsystems can be found. These systems record all steps of processes in execution. Based on this data, process mining aims at discovering process knowledge. Moreover, it allows for obtaining both qualitative and quantitative insights into these processes (e.g. performance information). As factual execution data is used, an objective view is provided on how healthcare processes are really executed. Obviously, there is a clear difference between process mining and more traditional ways of investigating business processes. For example, by conducting interviews there is always the risk that answers are subjective, idealized, and give an incomplete view on reality.

As shown on the right side of the above figure, there are three types of process mining.

Perhaps the most popular process mining type is discovery. Here the focus is on discovering process models that effectively describe the process information that have been recorded in the hospitals' information system. For example, one kind of model may describe the typical steps taken before surgery. Another kind of model may focus on the collaboration between nurses and doctors in order to treat patients. Note that also models focusing on performance and data may be discovered.

Another important process mining type is conformance. Here it is checked if observed behavior in the event log conforms to a given model. For example, it may be checked whether a medical guideline which states that always a lab test and an X-ray needs to be done is always followed.

Finally, there is the extension process mining type. Here, information extracted from the log is projected onto the model. For example, performance information may be projected on a discovered healthcare process in order to see for which examinations a long waiting times exist.