Nuclear Techniques ›› 2017, Vol. 40 ›› Issue (2): 20404-020404.

• NUCLEAR ELECTRONICS AND INSTRUMENTATION •

### Algorithm optimization of MLEM in coded aperture imaging system

LI Hanping, WANG Feng, AI Xianyun

1. State Key Laboratory of NBC Protection for Civilian, Research Institute of Chemical Defense, Beijing 102205, China
• Received:2016-11-20 Revised:2016-12-14 Online:2017-02-10 Published:2017-01-24

Abstract:

Background: In gamma-ray imager, reconstruction algorithm directly affects the quality of the reconstructed image. The maximum likelihood expectation maximization (MLEM) iteration algorithm is widely used in modified uniformly redundant arrays (MURA) coded aperture for the reason of satisfactory performance of suppressing noise, improving signal noise ratio (SNR) and reducing distortion. But MLEM iteration algorithm also has shortcomings, like amplifies the noise with the increase of the iterative times. Purpose: This study aims to improve the ability of suppressing noise for MLEM iteration algorithm, and increase precision and resolution of coded aperture imaging system. Methods: Based on the de-noising method of complementary coded aperture, a correction factor α for MLEM algorithm modification is proposed to control the convergence rate of the complementary MLEM iteration algorithm. Both Monte Carlo simulation and experimental date are used to verify the effectiveness of the complementary MLEM iteration algorithm. Results: The results of Monte Carlo simulation and experiment prove the complementary MLEM iteration algorithm is efficient and practicable in coded aperture image. Its efficiency is related to the modifying factor α. The fitting formula of α-μt in the complementary MLEM iteration algorithm is α=-0.6+0.12421μt-0.01252(μt)2. Conclusions: The coded aperture gamma camera is capable of imaging gamma-rays instantly. It can form radioactive 2D-map of different radionuclides. The reconstruction algorithm is important for coded aperture gamma camera. With the development of coded aperture, more and more study of algorithm modification will focus on it.

CLC Number:

• TL812