Restoration of Tc-99m methyl diphosphonate bone scan image using Richardson-Lucy algorithm

Area of Science:

• Nuclear Medicine Imaging
• Image Processing and Restoration

Background:

• Bone scintigraphy using Technetium-99m methyl diphosphonate (Tc-99m MDP) is crucial for diagnosing various bone pathologies.
• Image quality limitations in Tc-99m MDP scans can affect diagnostic accuracy.
• Image restoration techniques can potentially improve the diagnostic value of scintigraphic images.

Purpose of the Study:

• To determine the optimal input parameters for the Richardson-Lucy algorithm to restore Tc-99m MDP whole-body bone scan images.
• To objectively and subjectively evaluate the improvement in image quality after restoration.

Main Methods:

• Experimentally determined optimal point spread function (PSF) and number of iterations for the Richardson-Lucy algorithm.
• Utilized 60 Tc-99m MDP whole-body bone scan images for parameter estimation and validation.
• Assessed image quality using objective metrics (BRISQUE, MB, DE, EBCM) and subjective evaluation by nuclear medicine physicians.

Main Results:

• Restored images demonstrated significantly improved sharpness, reduced brightness, enhanced detail, and less edge contrast.
• Objective metrics indicated superior image quality in restored images.
• Nuclear medicine physicians assigned an average quality score of 4.00 to restored images, indicating significant visual improvement.
• Optimal PSF parameters were identified as size = 11 pixels and sigma = 1.75 pixels, with 10 iterations.

Conclusions:

• The Richardson-Lucy algorithm effectively restores Tc-99m MDP bone scan images, enhancing overall image quality.
• The determined optimal PSF parameters (size: 11 pixels, sigma: 1.75 pixels) are crucial for effective image restoration.
• Optimized image restoration can improve the diagnostic utility of bone scintigraphy by enhancing lesion detectability.