Subpopulation process mining based on graph matching
Type: Bachelor CS
If you are interested please contact :
Every consult, scan, measurement or treatment is an event in a patient's care path. Viewing from this perspective, a care path is like a business process and can be studied from this same perspective. "Process mining is a family of techniques in the field of process management that support the analysis of business processes based on event logs" (Wikipedia: https://en.wikipedia.org/wiki/Process_mining). As  says: "Processes in organisations such as hospitals, may deviate from intended standard processes, due to unforeseeable events and the complexity of the organisation. For hospitals, the knowledge of actual pa- tient streams for patient populations (e.g., severe or non-severe cases) is important for quality control and improvement. Process discovery from event data in electronic health records can shed light on the patient flows, but their comparison for different populations is cumbersome and time-consuming."
The mentioned paper  proposes to compare processes for different patient subpopulations by several means among others by graph similarity. This is assignment is meant to continue on this path: come up with more graph similarity measures and compare their results for some scenario/data set.
The graph similarity measures of  are only first trials. We have been comparing subpopulations of breast cancer patients based on their age, which malignacy score they had when they were first examined, and whether they were referred by the GP or some breast screening program. We expect that there could be better graph similarity measures for this purpose.