Textural Analysis Based Classification of Digital X-ray Images for Dental Caries Diagnosis

Abstract

In this paper, we propose a suitable textural feature for diagnosis of dental caries in digital radiographs. The dental diagnosis system consists of Laplacian filter for image sharpening, adaptive threshold and morphological operations for segmentation, and support vector machine (SVM) as a classifier. In segmented image, textural features are extracted, and applied to the classifier, to classify the image as caries or normal. Experimental results indicate that GLCM (Grey Level Co-occurrence Matrix) and GLDM (Grey Level Difference Method) textural features are giving better performance measures as compared to other types of textural features with an accuracy of 96.88%, sensitivity of 1, specificity of 0.8667 and precision of 96.08%. The data were analyzed by Analysis of Variance (ANOVA), at a significant level of 5%. This result indicates that the interaction of feature extraction methods on performance measures are significant. Hence, GLCM or GLDM features provide reliable decision support for dental caries diagnosis.

Revolutionizing Dental Caries Diagnosis through Artificial Intelligence

“…Techniques like gray-level co-occurrence matrix (GLCM), local binary patterns (LBP), and wavelet transforms are employed to extract texture-related features. These methods quantify variations in the image, capturing nuances that can signal the presence of early or advanced carious lesions [38].…”