Dental X-ray Image Enhancement using Evolutionary Algorithms

Dental x-ray radiography is often done with low radiation dosage which results in a slightly dark, low in contrast and noisy x-ray image. The quality of dental radiograph is low due to the amount of x-ray radiation directed is less on purpose to reduce the risk of various health issues. To solve these problems, this paper presents quantitative and qualitative analyses between the three most widely used evolutionary techniques, namely Particle Swarm Optimization (PSO), Cuckoo Search Algorithm (CSA), and Artificial Bee Colony Optimization (ABC). In this research paper, we have evaluated the above-said techniques on periapical dental x-ray images. The quantitative analysis of obtained results is done using quality metrics such as mean signal error (MSE), peak signal to noise ratio (PSNR), and structure similarity index (SSIM). Qualitative analysis was done with the help of a dental expert. The dentist evaluated the quality of results based on the application and diagnostics required in the periapical dental x-ray pathology. Experimental findings show that implemented evolutionary algorithms managed to enhance the pathology better than the original image. Results obtained show that image enhancement achieved using PSO and ABC algorithms is better than CSA algorithm.