Artifacts in slab average-intensity-projection images reformatted from JPEG 2000 compressed thin-section abdominal CT...

Area of Science:

• Medical Imaging
• Radiology
• Image Compression

Background:

• Thin-section CT images offer high detail but result in large data sizes.
• Image compression is crucial for efficient storage and transmission of medical imaging data.
• Assessing the impact of compression on image quality is vital for maintaining diagnostic accuracy.

Purpose of the Study:

• To evaluate the effect of Joint Photographic Experts Group (JPEG) 2000 compression on thin-section abdominal CT images.
• To compare compression artifacts in thin sections, thick sections, and average intensity projection (AIP) images.

Main Methods:

• Twenty thin-section (0.67 mm) abdominal CT datasets were compressed using JPEG 2000 at various ratios (4:1 to 15:1).
• Artifacts were quantified using Peak Signal-to-Noise Ratio (PSNR) and High Dynamic Range Visual Difference Predictor (HDR-VDP).
• Radiologists graded artifacts, and statistical analyses (Friedman test, exact tests) were performed.

Main Results:

• Compression artifacts increased with higher compression ratios across all image types.
• Artifacts were significantly lower in AIP images compared to thin and thick-section images (p < 0.0001).
• The visually lossless compression threshold decreased in the order of AIP, thick-section, and thin-section images.

Conclusions:

• Compression artifacts in thin-section CT are substantially reduced when reformatted into AIP images.
• Compressing thin-section images to levels acceptable for thick-section images is justifiable when using AIP review techniques.