Gingivitis detection by fractional Fourier entropy with optimization of hidden neurons

Abstract

Oral health care has always been a more expensive medical program than other types of treatment. Gingivitis is one of the oral problems that often be overlooked until the pain of gum is noticed by patients. When initial symptoms of gingivitis occur, people are more likely to ignore mild symptoms and choose not to treat them. Until they feel great pain, they have to go to the hospital for treatment, which is too late for doctors and patients. The complexity of operation will increase and may lead to serious life risks when the disease becomes worsen, therefore it is important for dentists to detect the underlying warning signs of gingivitis as soon as possible. In order to reduce the difficulty of artificial detection and improve the detection efficiency, we propose a new intelligent detection method for gingivitis. The key eigenvalue extraction of the gingivitis image in this approach is derived from the fractional Fourier entropy (FRFE) feature extraction method. This efficient eigenvector value is used for a standard genetic algorithm (SGA) classification together with an optimization of hidden neurons (OHN), which allows us to obtain a more efficient detection result than other gingivitis recognition systems.