Detection of Pulpitis Using MFCC and CNN1D

¹ School of Electrical Engineering, Telkom University, 40257 Bandung, Indonesia
² School of Dentistry, Universitas Padjadjaran, 40132 Bandung, Indonesia

^* Corresponding author’s email: chandrasyab1@gmail.com

Abstract

In this paper, we present a crucial problem the public faces in maintaining dental health, specifically related to pulpitis. Pulpitis is an inflammation of the dental pulp tissue caused by various factors such as bacterial infection, trauma to the tooth, or tooth decay. We responded to this challenge by creating an innovative solution to detect and distinguish pulpitis from healthy teeth. This solution will help dental professionals diagnose and treat pulpitis more effectively. The method we applied in this research is pulpitis detection using audio signals with machine learning algorithms. In this study, we used a CNN1D model with the addition of MFCC as a feature extraction with the hyperparameters Adam optimizer, learning rate 0.001, batch size 32, and test size 0.2. The model evaluation used a confusion matrix to assess the model’s ability to predict based on sound. Implementing machine learning in pulpitis detection through audio signals can help health workers accurately diagnose the condition.