Title: User-centered Approaches for Feature Extraction in CT Scanned Data
Name: Christina Gillmann
Phone: 0631 – 205 - 2642
CT ( Computed Tomography ) scan data is a widely used imaging method in various application as mechanical engineering and medicine. Besides the raw CT scan, detected features can raise a deeper insight to the reviewed scan. Although various approaches for feature detection and reviewing in CT scans are available, most of them cannot directly be applied by the user as they usually lack in the background knowledge to understand and utilize the approaches correctly.
This project aims to invent new feature extraction and visualization approaches by inserting visual exploration and user guidance to the feature extraction process, thereby the user can perform and review an extraction process independent and control if the results are correct.
In order to achieve this, the feature extraction process will be designed as a new workflow that allows the user to perform segmentations, geometry extraction and simulation of the examined CT scan and use the generated data according to the underlying use case. Therefore visualization is used as a feedback loop to let the user understand the feature extraction output.
As a result this project will develop a workflow for user centered feature extraction in CT scans that can be applied to several areas. To show the applicability of the workflow, real world examples will demonstrate the performance of the workflow in different fields.
Left: Aimed Workflow. Right: Features of coronary vessels. a) CT of pig heart b) Segmentation
c) Boundary extraction d) Centerline e) Flow simulation