Convolutional Neural Networks Towards Diagnosis of Dermatosis
Kangyu Gao, Qingyong Zhang, Haoran Wang · Journal of Physics Conference Series · 2019
In the field of machine learning and image recognition, deep convolutional neural networks have become a powerful tool for solving practical problems. Based on convolutional neural network's fast speed and high accuracy, we applied it to the diagnosis of pigmented skin in the medical field. By training the models on the dataset of ISIC, we eventually achieved up to 93.1% accuracy rate. If applied as an actual auxiliary tool, the diagnostic accuracy of related illness will be greatly improved.