History of MITK

MITK 1.x: Implemented basic encapsulation of data structures and algorithms, which included most fundamental functions of data handling, filtering, segmentation, registration and visualization.

MITK 2.0: The Out-of-Core mechanism was introduced, which supported the buffering of large datasets.

MITK 2.2: Some reconstruction functions were introduced which formed a new branch of the algorithm hierarchy.

MITK 2.3: The project was reorganized with the CMake tool provided by Kitware. The source code as well as the binaries were divided into multiple modules, which simplified the management, and made it possible that more compiling environments were supported.

Currently, MITK is extending itself in multiple aspects including the support of GPU alogorithms in a more flexible framework.

MITK 2.4.0 Key changes from 2.3.6:

(1) NEW Additional\mitkOpenCL module, this module is a GPU-computing sub-framework of MITK based on OpenCL. Currently, it contains some classes for volume rendering. Comparing to CUDA based volume rendering, this module is designed on a more fined-grained basis thanks to the dynamic-compilation mechanism of OpenCL.

(2) The OpenCL classes in mitkVisualization2 introduced during the last release are removed for compatibility issues.

(3) NEW volume thinning algorithms for centerline extraction are added in mitkCommon.

(4) NEW mitkVis2SurfaceScene scene in mitkVisualization2 added providing depth-peeling feature.

MITK 2.3.6 Key changes from 2.3.5:

(1) NEW mitkIsosurfaceCurvatureFilter in mitkCommon: A volume filter class for calculation of isosurface curvature from volume datasets.

(2) NEW mitkPath in mitkCommon: for storage of paths such as center lines.

(3) NEW mitkCLMIPVRScene in mitkVisualization2: OpenCL based MIP Volume Rendering.

(4) NEW mitkCLIsosurfaceScene in mitkVisualization2: OpenCL based Isosurface Volume Rendering.

(5) NEW mitkCLFullVRScene in mitkVisualization2: OpenCL based Full Volume Rendering.

(6) NEW mitkCLEndoscopySliceScene in mitkVisualization2: OpenCL based endoscopy navigation.

(7) NEW mitkCUDADeltaAccStateVRScene in mitkCUDAVolumeRendering: A state-transition model based volume rendering scene, implemented using CUDA.

(8) NEW mitkCUDAValueBasedStateVRScene in mitkCUDAVolumeRendering: Another state-transition model based volume rendering scene, implemented using CUDA.

(9) NEW mitkCUDAWallCarvingVRScene in mitkCUDAVolumeRendering: A endoscopy volume-rendering scene with Wall-Caring.

(10) NEW mitkCUDAWallPenetrateVRScene in mitkCUDAVolumeRendering: A endoscopy volume-rendering scene with Wall-Penetrating.

(11) Centerline navigation support for endoscopy volume rendering scenes.

  1  2  3  4  5  6  7  8  9  10  next

Address:No.95, Zhongguancun Road, Haidian District, Beijing City, China

Copyright (C) 1996-2013 - The Intelligent Medical Reasearch Center, CASIA - All Rights Reserved