Image-quality figure evaluator based on contrast-detail phantom in radiography

Area of Science:

• Medical Imaging
• Radiology
• Image Processing

Background:

• Accurate diagnostic information and minimal radiation dose are crucial in radiology.
• Contrast-detail phantoms are used to assess digital X-ray imaging system quality and radiation dose.
• Manual evaluation of phantom images by radiologists is subjective, time-consuming, and inconsistent.

Purpose of the Study:

• To develop an automated, objective, and efficient method for evaluating image quality figures (IQF) in digital X-ray imaging.
• To improve upon the accuracy and consistency of traditional manual evaluation methods.

Main Methods:

• An image processing-based IQF evaluator (IP-IQF) was developed to automatically measure phantom image quality.
• Test images were processed by separating the phantom into squares, stretching contrast, and analyzing regions using gradient and run-length features.
• A genetic algorithm was employed to optimize the parameters of the IP-IQF evaluator.

Main Results:

• The IP-IQF evaluator demonstrated efficient measurement of phantom image quality.
• Experimental results showed the IP-IQF evaluator to be more effective than both radiologists and the CoCIQ program in IQF evaluation.

Conclusions:

• The proposed automated IQF evaluator is more sensitive and accurate than manual radiologist observation and the CoCIQ computer program.
• A genetic algorithm effectively determined optimal parameters for the IQF evaluator, enhancing its performance.