ANN-ANT-LION-MLP Ensemble Transfer Learning Based Classifier for Detection and Classification of Oral Disease Severity

Deep learning is subset of artificial intelligence which is considered as a backbone for object identification in computer vision technology. For societal issue deep learning based intelligent system is a significant approach to solve problem of society prominently in healthcare becomes a debatable concern. Medical image processing is emerging part of Al application in healthcare. In conventional Method of oral disease identification of severity rely upon Medical imaging reports and sometimes symptoms may vary from actual disease. That’s is why automation can be give a complimentary prediction option to Dentist and this can cake medical report more accurate. Thus, in this paper we have utilized same approach of using deep learning for oral disease severity detection and classification. Oral disease in India is a major problem of endanger lives of millions with severe cases like oral cancer. We have done a survey for overall cases of oral being addressed in India for giving a backbone of deep learning for society. Moreover severity detection in oral cavity through deep learning requires algorithm used to be more precise in terms of precision and accuracy. For that we have optimized convolution neural network using bio inspired Meta heuristic algorithm (antlion and dragon fly optimizer) which give 89% accuracy by testing optimization cases of testing threshold of ROI feature extraction. Also we have used transfer learning model based on multilayer perceptron (MLP) for training and testing our multiclass neural model with the accuracy of94.6%.