Three-dimensional Visualization of Medical Images and Signals

Medical images produced by Magnetic Resonance Imaging (MRI) and Computed Tomography (CT) can be visualized in 3D by applying computer graphics. In this project, several software was developed to display segmented images, raw images, and multimodal images in which simulated electric signals and patient images were combined. 

Imaging Software (Sun/Xview & UNIX/NeXTstep)

Imaging software [1,2] was developed to Sun/Xview environment during the spring 1994. Thereafter the software was developed further and ported to NeXTstep environment. Imaging software enables efficient visualization of contour segmented images (one contour per slice / tissue). The projection plane can be selected arbitrary and simulated electric signals can be visualized using different color maps. The user interface of the software is presented in Fig 1. The implementation of the Imaging software is based on heuristic triangulation, axonometric projection, Z-buffering, and Phong shading.

Figure 1: The user interface of the Imaging software running in NeXTstep environment.

Medimag Software (PC/Windows95)

Medimag software [3] was developed to PC/Windows95 environment using C++ language. It is capable of displaying voxel-based images, either segmented or raw. Furthermore, the software allows volume cuts, definition of arbitrary projection plane, and simultaneous visualization of numerous 3D structures. The user interface of Medimag is presented in Fig 2. The implementation is based on 3D flood filling, polygon surfaces, center of gravity, and Phong shading.

Figure 2: The user interface of Medimag software running in Windows environment. The software allows visualization of raw or segmented medical images.

Also Medimag is capable of presenting measured or simulated electric signals on a 3D surface by applying colors. In addition, the effect of transparency and arbitrary volume cuts enable efficient visualization of medical images. The software operated relatively fast, requiring few seconds per projection (Pentium 200MHz). Some images generated using Medimag are presented in Fig 3.

Figure 3: 3D images produced using Medimag software; (AB) Scalp and skull from a raw CT image set, (C) segmented nasal cavities and paranasal sinuses from a CT image set, (D) combination of raw and segmented MR images in which the scalp is partially transparent, (E) combination of raw and segmented MRI, (F) a combination of segmented MRI and a simulated signal.

Pilot Studies

The imaging software was utilized in the visualization simulated EEG signals of epileptic patients. The Medimag software has been applied in numerous research projects including visualization of neural structures and nasal structures (Fig 3C, 4), etc. Also pathological conditions, such as tumors and lesions have been displayed. The resulting images were high quality. Due to numerous features of the software, very accurate and intuitive images can be presented.

Figure 4: 3D reconstruction of nasal structures; Nasal cavities and paranasal sinuses are presented together with bony structures. The skull is partially transparent.

REFERENCES

[1] Heinonen, T. & Eskola, H. 1995. Reconstruction of Three-Dimensional Surface of the Brain Using Triangular Facet Surface and Heuristic Methods. Tampere University of Technolgy, Ragnar Granit Institute, Report. 9. Tampere. 4, 17 p.

[2] Heinonen T. 1994. Three-Dimensional Presentation of Medical Images and Signals, MScThesis, Tampere University of Technology, 101 p.

[3] Heinonen T, Visala K, Blomqvist M, Eskola H, Frey H. 1998. 3D Visualization library for multimodal medical images, Computerized Medical Imaging and Graphics, 22(4): 6 p. 

For further information, please contact:

Tomi Heinonen, Email: tomi@ee.tut.fi, fax: +358 3 247 4013, tel: +358 40 547 1451