| |
 |
| Endoscope Distortion Correction in (Texture Classification-based) Automated Diagnosis Support Systems (Undistort)
|
| |
Project Description:
Undistort (FWF project 24366) is a project devoted to the investigation if knowledge about the type and
extent of optical distortion in endoscopes can be exploited to improve the accuracy of texture-classification based automated
diagnosis support systems. In particular, we focus on the gastrointestinal tract. We rely on datasets generated in earlier
projects with
Michael Häfner from the St. Elisabeth Hospital Vienna (colon data, aimed at polyp classification and cancer detection), and with Andreas Vecsei from the Vienna St. Anna Childrens Hospital (duodenal data, aimed at diagnosis of celiac disease).
|
|