An Optimized Predictive Model Based on Deep Neural Network for Detection of Skin Cancer and Oral Cancer

Abstract

Cancer is becoming a serious illness having a high mortality range. Cancer is a disease that causes death even before the age of 60, according to recent research conducted in 90 nations. More than 9 million individuals will pass away in 2020, with cancer accounting for 1 in 6 of those fatalities. Skin and oral malignancies are considered the greatest prevalent cancer types. Predicting the cancerous type based on pictures is crucial. This research work introduces revolutionary deep-learning methodologies for skin and oral cancer diagnosis at the start of the disease. The main goal of this research work is to assist physicians in early oral and skin cancer diagnosis. Convolutional neural networks and VGG19 are combined in the proposed model to identify and classify various forms of skin and oral cancer. The author pares the dataset with other competing classifiers to validate the algorithm. The findings demonstrate that these methodologies’ hybridization provides greater accuracy than conventional techniques. The proposed model has a 97.17% accuracy, a 92.19% precision, a 92.18% recall, and a 93.10% F1-score value. In the future, this technique can be improved with additional features for treating other tumors.