Sgfin.github.io Website Review

harvard medical school, mit i am a md-phd candidate working with dr. isaac kohane. my research interests center on the intersection of machine learning and healthcare. my current projects focus on exploring applications of de*p learning to medical images, medical text, and various 'omics. harvard medical school, mit md-phd program, mstp fellowship ba, human biology - biocomputation research mentor: daniel rubin. towards generative adversarial networks as a new paradigm for radiology education. medical students and radiology trainees typically view thousands of images in order to ”train their eye” to detect the subtle visual patterns necessary for diagnosis. nevertheless, infrastructural and legal constraints often make it difficult to access and quickly query an abundance of images with a user-specified feature set. in this paper, we use a conditional generative adversarial network (gan) to synthesize 1024 × 1024 pixel pelvic radiographs that can be queried with conditioning on fracture status. we demonstrate that the conditional gan learns features that distinguish fractures from non-fractures by training a convolutional neural network exclusively on images sampled from the gan and achieving an auc of > 0.95 on a held-out set of real images. we conduct additional an*lysis of the images sampled from the gan and describe ongoing work to validate educational efficacy. generative adversarial networks (gans) are a very cool technique that uses dualing neural networks to try to learn to approximate a data distribution. they have been used to create some beautiful images, but some researchers have current techniques don't appear to learn the full data distribution they're trying to approximate. as a researcher, this makes me nervous about using gan-sampled images as an data substitutes in high-stakes clinical applications, but i'm still very interested in considering ways we could find productive uses for gans while we're figuring out how to solve their limitations. as a medical student on a radiology rotation, i spent many hours