Research Area:  Artificial Intelligence
This thesis is composed of four chapters in neuroscience, psychiatry, drug screening, and dermatology all linked by the common thread of using AI to enhance the expert.We go on to make a case for the scalability of neuroscience, adopting a paradigm focused on the recording of in-vivo information and the reporting of that information to external sources.
Sychiatry, we demonstrate that an AI system composed of a camera and eye-tracker can classify Fragile-X-Syndrome Autism against generic developmental disorders directly from a video stream,by monitoring the interaction between a patient and physician.
Drug Screening focuses on cellular biology, where we build a computer vision system capable of predicting molecular stain protocols of neurons in petridishes, imaged via bright-field and confocal microscopy.
ermatology, we apply computer vision to the classification, detection, and track-ing of skin cancers in clinical and dermoscopic photographs. We begin by demonstrating that a deep neural network can achieve the performance of board-certified dermatologists at classifying skin cancers against benign neoplasms.
Name of the Researcher:  Andre Esteva
Name of the Supervisor(s):  Sebastian Thrun
Year of Completion:  2018
University:  Stanford University
Thesis Link:   Home Page Url