Dr. Karen Schroeder is a postdoctoral research scientist working with Dr. Mark Churchland and Dr. Larry Abbott at Columbia University’s Zuckerman Institute. She is working to create brain-machine interface systems that accurately decode a wide range of complex and naturalistic voluntary movements. She received her Ph.D. in biomedical engineering in 2016 under Dr. Cynthia Chestek at the University of Michigan, where she developed decoding techniques and hardware for dexterous finger brain-machine interfaces. After moving to Columbia, she began working on a new class of movement decoding strategies that leverage reliable population-level latent features in multiunit recordings. She applied these strategies to decode rhythmic arm movements for the control of virtual ego-motion. Her approach combines novel statistical and computational methods with closed-loop experimental testing to develop and refine decoding techniques. Her current research focuses on understanding the structure of population activity in motor cortex across multiple kinds of behaviors, and using that knowledge to create high performance decoders that can capably control prosthetic devices.