On the use of Deep Learning Approach for Individualized Treatment

Since the beginning, progress in medicine has been strongly linked to technological advances, and more recently many discoveries and advances in medicine and pharmacology are dependent on new tools and techniques from computer science. In particular, with the revival of artificial intelligence and the re-emergence of machine learning through deep neural networks, many researchers are proposing to incorporate machine learning techniques into systems for disease prediction, diagnosis, or drug design, as part of the individualized medicine that is emerging as a promising way to treat patients. In this work, we focus on personalized medicine through individualized therapy, based on treatment outcomes. For that purpose, we use a convolutional neural network, as a deep learning algorithm known for its effectiveness in features extraction from a given pattern. In our case, a pattern is a collection of variables that describe the treatment and the classes stand for the outcomes of the treatment. Experiments are carried on a public dataset that describes the wart treatment and its results. Experimental results show the potential of the CNN to predict the correct outcome, and hence its benefits in the context of personalized medicine.