[Study and realization of multidimensional visualization techniques for multimodality medical images].

Area of Science:

• Medical Imaging
• Computer-Aided Diagnosis
• Scientific Visualization

Context:

• Current medical imaging modalities like CT, MR, PET, and SPECT generate large datasets.
• Efficient visualization tools are crucial for accurate clinical interpretation and diagnosis.
• Existing visualization systems may lack multimodality support or advanced rendering techniques.

Purpose:

• To design and implement a multimodal medical image visualization system.
• To integrate various visualization techniques including multi-planar reconstruction (MPR), curved planar reformation (CPR), direct volume rendering (DVR), indirect volume rendering (IVR), and maximum intensity projection (MIP).
• To evaluate the system's performance and clinical utility on diverse medical imaging data.

Summary:

• A novel visualization system was developed using open-source VTK on the VS2008 platform.
• The system supports visualization of Computed Tomography (CT), Magnetic Resonance (MR), Positron Emission Tomography (PET), and Single-Photon Emission Computed Tomography (SPECT) data.
• Multiple advanced visualization methods are implemented, enabling comprehensive analysis of medical images.

Impact:

• Clinical practice indicates the system provides stable performance and convenient image reading across different modalities.
• The system's capability to reconstruct over 2,000 CT slices meets clinical requirements for speed and quality.
• Enhanced visualization facilitates more accurate and efficient diagnosis in clinical settings.