With the establishment of the clinical state cancer registries according to §65c SGB V and the introduction of the nationwide ADT-GEKID basic data set, data on the treatment and course of cancer are now also collected. Due to their complexity, the evaluation of these data can currently only be implemented with great, disproportionate manual effort without suitable procedures and tools. In the SePaMiM project, we would like to investigate to what extent the analysis of complex treatment and disease courses from registry data can be supported with the help of suitable IT systems.
SePaMiM aims to support epidemiologists in clinical cancer registries in the analysis of disease and treatment course data by means of suitable IT systems.
Two primary approaches exist for analysing sequential data. The first aims to look for sequences in the history data that match a given pattern. The second approach comes from data mining and deals with algorithms for finding interesting or frequently occurring patterns in the sequence data. In the SePaMiM project, both approaches will be pursued.
Firstly, the staff of the clinical cancer registries are to be supported in the selection of patient cohorts on the basis of specific treatment histories. This should enable the staff to answer concrete questions, for example in the context of a quality conference.
Based on this, artificial intelligence methods, especially from the field of data mining, are then used to automatically discover patterns in the course data. In the field of data mining, sequential pattern mining, for example, deals with algorithms for finding interesting or frequently occurring patterns in sequence data.