: This paper describes a method for the enhancement of curvilinear structures such as vessels and bronchi in 3D medical images. A 3D line enhancement filter is developed with the aim of discriminating line structures from other structures and recovering line structures of various widths. The 3D line filter is based on a combination of the eigenvalues of the 3D Hessian matrix. Multi-scale integration is formulated by taking the maximum among single-scale filter responses, and its characteristics are examined to derive criteria for the selection of parameters in the formulation. The resultant multi-scale line-filtered images provide significantly improved segmentation and visualization of curvilinear structures. The usefulness of the method is demonstrated by the segmentation and visualization of brain vessels from MRI (magnetic resonance imaging) and MRA (magnetic resonance angiography), bronchi from a chest CT, and liver vessels (portal veins) from an abdominal CT. Keywords: 3D image ...

This paper describes 3D image filters for the enhancement of specific local intensity structures such as line and sheet, and its application to tissue classification for volume rendering. Multi-channel classification is performed by combining different 3D image filter outputs. The resulted method significantly enlarges the scope of volume rendering, especially in the medical domain. We show the usefulness of the method for different visualization problems. 1 Introduction Volume rendering is a powerful visualization tool especially for medical application [1],[2],[3],[4]. Basic requirement in medical application is to visualize specific tissues of interest with the relation to surrounding structures. Tissue classification is one of the most important processes in the volume rendering pipeline. Most commonly, this process is done based on the histogram of intensity values in original 3D images. Probabilistic or fuzzy classification has been used instead of binary classification in orde...

In order to clearly depict densitometric as well as geometric information in vascular visualization from 3D data such as MR and CT angiographies, a new visualization method called local maximum intensity projection (LMIP) is proposed. LMIP is an extended version of MIP (maximum intensity projection). LMIP differs from MIP in that the latter method selects the maximum value along an optical ray, whereas LMIP selects the first local maximum value encountered that is larger than a pre-selected threshold value along an optical ray from the viewpoint in the viewing direction. Examples are presented in which LMIP is used to visualize renal vessels from CT angiography data and cerebral vessels in the vicinity of an aneurysm from phase-contrast MR angiography data. It is demonstrated that LMIP can clearly depict geometric information, like shaded surface display (SSD) does, and densitometric information, as is done by volume rendering (VR), in a straightforward and objective manner. Keywords:...

In order to clearly depict densitometric as well as geometric information in vascular visualization from 3D data such as MR and CT angiographies, a new visualization method called local maximum intensity projection (LMIP) is proposed. LMIP is an extended version of MIP (maximum intensity projection). LMIP differs from MIP in that the latter method selects the maximum value along an optical ray, whereas LMIP selects the first local maximum value encountered that is larger than a pre-selected threshold value along an optical ray from the viewpoint in the viewing direction. Examples are presented in which LMIP is used to visualize renal vessels from CT angiography data and cerebral vessels in the vicinity of an aneurysm from phase-contrast MR angiography data. It is demonstrated that LMIP can clearly depict geometric information, like shaded surface display (SSD) does, and densitometric information, as is done by volume rendering (VR), in a straightforward and objective manner. Keywords:...