Comparative evaluation of JPEG and JPEG2000 compression in quantitative digital subtraction radiography

Area of Science:

• Radiology
• Medical Imaging
• Image Compression

Background:

• Quantitative digital subtraction radiography (DSR) is crucial for assessing bone changes.
• Image compression is necessary for efficient storage and transmission of DSR data.
• Evaluating the impact of different compression standards on DSR accuracy is essential.

Purpose of the Study:

• To compare JPEG and JPEG2000 compression effects on quantitative DSR.
• To determine acceptable JPEG2000 compression ratios (CR) for DSR.

Main Methods:

• In vitro study using pig mandible sections.
• Images acquired at 'Baseline', 'No change', and 'Gain' (simulated bone chips).
• Images compressed using JPEG and JPEG2000 at CRs 1:7, 1:16, 1:22, 1:31; distortion assessed via pixel error and PSNR.

Main Results:

• JPEG showed less distortion at low CRs; JPEG2000 superior at higher CRs.
• No significant difference between JPEG and JPEG2000 at CR 1:7.
• JPEG2000 better for 'no change' measurements at higher CRs, but bone gain volumes were underestimated.
• JPEG overestimated bone gain volumes.

Conclusions:

• JPEG and JPEG2000 perform similarly at CR 1:7 for DSR, provided images are registered before compression.
• JPEG2000 is superior to JPEG at higher CRs but introduces distortions too high for reliable quantitative DSR.