Topic 13: Level sets¶

1. Read and discuss the paper by Li et al..
2. Generate a small test-object in a 3D image (e.g. a simple sphere).
3. Implement the Local Binary Fitting (LFB) Levelset algorithm and test it on this test-object.
4. Download the liver-imaging dataset from Brightspace (see the data for assignments-tab). This dataset contains a series of 3D volumes in Dicom format. In particular, the liver is imaged while contrast has been injected.
5. Segment the liver using your levelset implementation from e.g. the last volume in this dataset. In this volume the contrast is highest of the series. Even higher contrast can be obtained if you subtract the first volume from the last one. To that end you might ask your colleagues from Topic 9 to supply you with a registered series (they work on the same data).
6. Discuss the results.

References

1. Chunming Li， Kao Chiu-Yen， Gore John C.，et al. Implicit Active Contours Driven by Local Binary Fitting Energy ［C］. IEEE Conference on Computer Vision and Pattern Recognition， 2007，（1）：1-7.