Identifying the incidence level of periodontal disease through hyperspectral imaging

摘要

This study proposes a method for staging periodontal disease through hyperspectral imaging technique. The study is composed of two parts. The first part shows the spectrum of gingival images that is simulated by hyperspectral imaging systematic analysis. The second part uses the principal component score chart for staging periodontal disease based on the obtained mean simulation spectrum. The experimental results show the staging results of periodontal disease through principal component score chart, allowing us to clearly identify the spectral features of normal gingiva and mild, moderate, and severe periodontitis. Our method can identify the stage of periodontal disease accurately and has a sensitivity and specificity of approximately 0.83 and 0.938, respectively.