Deep Learning Data Augmentation for Raman Spectroscopy Cancer Tissue Classification

Recently, Raman Spectroscopy (RS) has demonstrated to be a non-destructive way of cancer diagnosis, due to the uniqueness of RS measurements in revealing molecular biochemical changes between cancerous vs. normal tissues and cells. In order to design computational approaches for cancer detection, the quality and quantity of RS tissues are the basis for accurate prediction. In reality, however, obtaining skin cancer samples is difficult and expensive due to privacy and other constraints. With a small number of samples, the training of the classifier is difficult, and often results in overfitting. Therefore, it is important to have more samples to better train classifiers for accurate cancer tissue classification. To overcome these limitations, this paper presents a novel generative adversarial network based skin cancer tissue classification framework. Specifically, we design a data augmentation module that employs a generative adversarial network to generate synthetic RS samples in different classes. The original tissue samples and the generated data are merged to train classification modules. Experiments on real-world RS data demonstrate that generative adversarial network can be successfully used for data augmentation, in order to train accurate skin cancer tissue classifiers.