Penetration of molars teeth using carbon dioxide laser / Noorsyam Yusof, Mohamad Adha Mohamad Idin and Noorezal Atfyinna Mohd Napiah.

• Main Authors: Yusof, Noorsyam, Mohamad Idin, Mohamad Adha, Mohd Napiah, Noorezal Atfyinna
• Format: Research Reports
• Language: English
• Published: 2011
• Subjects: RK Dentistry; Oral and dental medicine. Pathology. Diseases
• Online Access: http://ir.uitm.edu.my/id/eprint/42030/1/42030.PDF; http://ir.uitm.edu.my/id/eprint/42030/
• Institution: Universiti Teknologi Mara

Drilling is the most common approach in dental treatment. It takes time and sometime causes a pain on patient. By drilling the teeth does not achieve the desire depth should be. So the new technology is welcome to solve this kind of problem. Different types of lasers had been used in dental applications since the early 1990’s for soft tissues (gums). Dental lasers have not been widely used because of their high cost and limited applications. A dental laser for preparation of cavities is a very new and potentially promising technology that will need several years of research and improvements. This research presents an artificial neural network (ANN) technique to predict and optimize the depth penetration of carbon dioxide laser used for teeth. This determination is important to dental treatment instead of using drill for drilling the teeth. The objective of this research is to determine the best parameter in carbon dioxide laser in order to achieve the desire depth penetration of teeth. The parameters are in terms of power intensity and exposure time. The combinations of this parameter are important to get the required depth penetration. Carbon dioxide laser is used to shoot the molars. The area that will expose to laser is enamel. The parameter setting is set differently according to the type of teeth. The collected data from the experiment will used to build the intelligence system. The model combines single hidden layer multilayer perceptron Artificial Neural Networks (ANN) for prediction and optimization. In this method, training and prediction performance of different ANN architectures are initially tested and the architecture with the best performance is further used for optimization. Finally, the best ANN architecture is found to show much better prediction capability compared to an experimental method in obtaining the depth penetration. Therefore, the new method is introduced in this research.