No-Reference-Based and Noise Level Evaluations of Cinematic Rendering in Bone Computed Tomography

Area of Science:

• Medical Imaging
• Radiology
• 3D Visualization

Background:

• Cinematic rendering (CR) is an emerging 3D post-processing technique for bone CT imaging.
• Volume rendering (VR) is a conventional method in diagnostic medical imaging.
• Evaluating CR's performance against VR in bone CT is crucial for clinical adoption.

Purpose of the Study:

• To assess the image quality and noise levels of CR compared to VR in bone CT.
• To quantify the performance differences using objective image quality metrics.
• To determine the potential of CR in enhancing diagnostic accuracy for bone pathologies.

Main Methods:

• Bone CT images of the face, shoulder, lumbar spine, and wrist were acquired.
• Cinematic rendering (CR) and Volume rendering (VR) were applied to the datasets.
• Blind image quality was evaluated using the Blind/Referenceless Image Spatial Quality Evaluator (BRISQUE).
• Noise levels were assessed using the Coefficient of Variation (COV).

Main Results:

• CR demonstrated superior image quality with lower average BRISQUE scores (39.87) compared to VR (46.44).
• The performance difference was more pronounced in bone CT images with metal artifacts.
• CR resulted in a 2.20-fold improvement in noise reduction (COV) compared to VR.

Conclusions:

• Cinematic rendering (CR) is a valuable tool for reconstructing high-quality 3D bone CT images.
• CR offers significant advantages in image quality and noise reduction over VR.
• CR holds promise for diverse applications within diagnostic medical imaging.