Grouping Similar Trajectories in Hospital Laboratory Data

Grouping of time series medical data is still a challenging task as it requires comparison of mutidimensional and temporally irregular data. In this paper, we attempt to overcome these problems by using multiscale comparison of directed trajectories. A set of time series containing different types of laboratory tests are mapped into directed trajectories representing the time course of patient states. Then the trajectories for individual patients are compared and grouped into similar cases. Experiments were conducted on the grouping tasks of artificially generated stroke trajectories of digits and ALB-PLT trajectories on chronic hepatitis patients. The results showed that the method could group the similar trajectories successfully.