Diagnosis of periodontal disease with Levenberg-Marquardt algorithm

Periodontal disease is one of the most common infectious diseases of the mouth. Correct and early diagnosis can reduce the incidence of adverse effects. The aim of this study was to evaluate the accuracy and efficiency of artificial neural network in the diagnosis of periodontal disease.
Material and The Diagnostic Study were performed from 230 periodontal disease cases in Zahedan Dentistry School in the period of time between 2015 and 2016. 5 variables including age, gender, plaque index, probing pocket depth, and clinical attachment loss index were evaluated in these people. The artificial neural network model with propagation algorithm of Levenberg-Marquardt training function was used. Positive predictive value and negative predictive value were used to evaluate the network at the test stage.
The results show that the back propagation network with the structure of 5-20-4-2 and the Levenberg-Marquardt algorithm and the use of the same transfer functions in all layers (sigmoid hyperbolic tangent) can be used as an efficient teaching function in Diagnosis of periodontal disease. The values of positive and negative predictive values were 94.7 and 80 percent respectively. The software output yielded significant amounts of time (4.5870) and regression for train, test and overall (0.7454, 0.9749, 0.9254).
It seems, artificial neural networks can be helpful for diagnosis of periodontal disease at least time.