Menoua Keshishian

PhD Candidate, Columbia University

About

I am a sixth-year Ph.D. candidate in Electrical Engineering at Columbia University and a member of the Neural Acoustic Processing Lab which is a part of the Zuckerman Mind Brain Behavior Institute. I am interested in the processing of speech and language in the human brain. Specifically, I use machine learning methods to study how the human auditory cortex analyzes the acoustic and linguistic content of speech as it is processed through the auditory pathway. Click here to see my full CV.

Below is a list of my projects and publications.

Research Projects

Image schematic of the research project
Linguistic encoding. Analyzing how different levels of linguistic content of speech are encoded in the neural representation of the listener's brain, using ridge-regression and feature permutations.
Image schematic of the research project
naplib-python. A python toolbox for analyzing neural-acoustic data such as electroencephalography (EEG) paired with acoustic stimuli.
Image schematic of the research project
PyTCI. A python toolbox to analyze the temporal integration windows of computational models (e.g., artificial neural networks) that process time-series input.
Image schematic of the research project
dSTRF. A python toolbox to analyze feed forward neural networks trained to predict biological neural responses to sound, by computing the locally linear operations they perform on the input.

Side Projects

Image schematic of the side project
CogTask. An application written in Rust for designing and running cognitive psychology tasks/experiments, with the goals of being low-latency and easy-to-use.

Publications

naplib-python: Neural Acoustic Data Processing and Analysis Tools in Python. Mischler, G, Raghavan, V, Keshishian, M, Mesgarani, N. Software Impacts (2023)

Joint, distributed and hierarchically organized encoding of linguistic features in the human auditory cortex. Keshishian, M, Akkol, S, Herrero, J, Bickel, S, Mehta, AD, Mesgarani, N. Nature Human Behaviour (2023)

Deep neural networks effectively model neural adaptation to changing background noise and suggest nonlinear noise filtering methods in auditory cortex. Mischler, G, Keshishian, M, Bickel, S, Mehta, AD, Mesgarani, N. NeuroImage (2023)

Understanding Adaptive, Multiscale Temporal Integration In Deep Speech Recognition Systems. Keshishian, M, Norman-Haignere, S, Mesgarani, N. Advances in Neural Information Processing Systems (2021)

Estimating and interpreting nonlinear receptive field of sensory neural responses with deep neural network models. Keshishian, M, Akbari, H, Khalighinejad, B, Herrero, JL, Mehta, AD, Mesgarani, N. eLife (2020)