Synthetic Group Difference Generator for Testing of Group Analysis Tools in DTI studies (2015)

Undergraduate: Matthew Leming

Faculty Advisor: Martin Styner
Department: Computer Science

---

Diffusion Tensor Imaging (DTI) is a tool often used in studies of white matter for comparison of the integrity of white matter across populations. Differences in white matter integrity are frequently indicative of the effects of drugs and mental disorders on neurobiology. A number of software tools have been designed in recent years, such as TBSS and FADTTS, that are able to detect and quantify average differences in measurements of DTI, in order to compare different populations of datasets. However, the effectiveness of these tools is often assessed by testing them on datasets of current subjects, whose group differences would be harder to assess. Synthetic Group Differences is a tool that deliberately creates average differences between groups of subjects by altering the principal eigenvalues of diffusion tensor datasets within a number of localized spheres across the dataset. In this way, group population tools can effectively be tested by the amount they detect such differences in the pre-defined dataset.