Knowledge Discovery and Machine Learning: Research in Gingivitis Detection

Gingivitis is a high-risk condition that causes dietary issues in older people. The study of gingivitis is more difficult in the realm of medical image analysis due to the absence and complexity of dental image analysis. As traditional clinical diagnosis takes time and money and necessitates a lot of physical effort on the part of competent clinicians. In contrast, deep learning allows for automated medicine via picture analysis. However, several obstacles remain in medicine, such as poor machine model performance, inadequate training data, and expensive labeling costs, to name a few. These difficulties encourage the development of data- and knowledge-aware deep learning approaches that can be used for a variety of medical activities without requiring considerable human labeling and that incorporate domain-specific information throughout the learning process. This paper reviews and analyses research in computer-aided diagnosis and medical image deep learning, with a focus on the challenges in the field of gingivitis image detection, and proposes model performance achieved by combining different image extraction methods and different classification methods. At the same time, some traditional feature extraction methods and standard computer-aided diagnosis methods are introduced. In this paper, a feature extraction model based on fractional Fourier entropy and wavelet energy entropy is proposed for gingival image segmentation, and various classification and optimization techniques are combined. By evaluating the reintegrated medical images, the experimental results of the gingivitis detection model based on fractional Fourier entropy feature extraction combined with particle swarm optimization neural network show that the detection method significantly reduces the detection space and the complexity of image information. The improved algorithm can cluster the sample data efficiently and accurately, and the accuracy is higher than that of advanced gingival image diagnosis technology.