A team of researchers is currently exploring how to use MRI images generated by artificial intelligence to train deep learning models designed to identify clinical abnormalities in imagining data, HealthITAnalytics reports. The research aims to overcome the challenge of obtaining abnormal images, which are uncommon, by employing a method to generate synthetic abnormal MRI images. In doing so, the researchers are able to create a computer-generated data set, which carries no privacy concerns, and could supplement data sets used for training AI. In a research paper, they state, “These results offer a potential solution to two of the largest challenges facing machine learning in medical imaging, namely the small incidence of pathological findings, and the restrictions around sharing of patient data.”
Full Story