Understanding Machine learning Applications in Cancer Prognosis and Early Detection

Abstract

Abstract: Machine learning and deep learning technologies have seen a recent growth in trend towards their applications in personalised and predictive medicine. The ML models are designed with an aim to supervise the progression of cancer within a patient and aid in its treatment. Cancer is quite a diverse condition, which means that an early diagnosis and timely screening is instrumental for treatment. An array of popular ML techniques is used in Cancer Prognosis including and not limited to Artificial Neural Networks, Decision Trees, Support Vector Machines, Bayesian Networks and other Deep Learning approaches. Each of these methodologies contribute to the development of predictive models. Each model developed is expected to substantially improve the accuracy of susception and recurrence prediction. However, a number of published studies that appear to have build over these models lack validation and/or appropriate testing. In this review, we analyse and present a view on recent development of ML approaches that are applied in Cancer Prognosis and Prediction modelling. Keywords: Machine Learning, Cancer Prognosis, Cancer Detection, Deep Learning, Artificial Intelligence