Level Set Techniques For Structural Inversion In Medical Imaging

Abstract

Most biological bodies are structured in the sense that they contain quite well-defined interfaces between regions of different types of tissue or anatomical material. Extracting structural information from medical or biological images has been an important research topic for a long time. Recently, much attention has been devoted to quite novel techniques for the direct recovery of structural information from physically measured data. These techniques differ from more traditional image processing and image segmentation techniques by the fact that they try to recover structured images not from already given pixel or voxelbased reconstructions (obtained, e.g., using traditional medical inversion techniques), but directly from the given raw data. This has the advantage that the final result is guaranteed to satisfy the imposed criteria of data fitness as well as those of the given structural prior information. The 'level-set-technique' [1][2][3] plays an important role in many of these novel structural inversion approaches, due to its capability of modeling topological changes during the typically iterative inversion process. In this text we will provide a brief introduction into some techniques that have been developed recently for solving structural inverse problems using a level set technique.

Level Set Methods for Structural Inversion and Image Reconstruction

“…This chapter can only give a rough overview of some techniques which have been discussed so far in the literature. For more details which go beyond the material covered here, the reader is referred to the recent review articles [20,[31][32][33]92], each of them providing a slightly different view on the topic and making available a rich set of additional references which the interested reader can follow for further consultation.…”

A high-order algorithm for multiple electromagnetic scattering in three dimensions

“…The above representation includes, but is not limited to, star shaped obstacles. Many interesting scattering geometries can be described in this way, including Gaussian random particles simulating ice crystals [32] (modeled from images obtained by a cloud particle imager instrument) and dust particles [37], erythrocytes [22,41], brain tumors [20], kidneys [6], all benchmark radar targets in [40], (after converting the material construction coordinates [40] to spherical coordinates [14]), and recently established three dimensional shape reconstruction surfaces in medical imaging and optical tomography [8,9,43,44].…”

A high-order tangential basis algorithm for electromagnetic scattering by curved surfaces

“…Many interesting and practically important scatterers can be described in this way, including Gaussian random particles simulating ice crystals (modeled from images obtained by a cloud particle imager) [30], dust particles [37], erythrocytes [22,40], brain tumors [21], all benchmark radar targets in [39], (after converting the material construction coordinates [39] to spherical coordinates [16]), and recently established three-dimensional shape reconstruction surfaces in medical imaging and optical tomography [12,11,41,42].…”

“…Scattering simulations also play an important role in medical sciences because scattering is an important tool, for example, in classifying bi-concave blood cells [22,40] and in image-guided neurosurgery using medical imaging, for example, of tumor models [12,21,41,42]. These obstacles are in general dielectric.…”