Characterizing Autism and Schizophrenia Using PRISM and Deep Learning
Name
Wu-juliawu-meng-eecs-2021-thesis.pdf
Description
Thesis PDF
Size
4.27 MB
Format
Adobe PDF
Checksum (MD5)
eddb506cdbf63e058e8637f59ba159eb
Author(s)
Wu, Julia
Advisor(s)
Bathe, Mark
Date Issued
June 2021
Publisher
Massachusetts Institute of Technology
Abstract
Schizophrenia and autism spectrum disorder (ASD) are two life-altering neurological diseases whose neurobiological bases are not yet well understood. This thesis explores the phenotypical expression of autism and schizophrenia at the synapse level by applying deep learning to multiplexed immunofluorescence data. Deep convolutional networks are developed and applied to analyze PRISM images of neurons treated with gene knockdown treatments corresponding to genes associated with autism and schizophrenia. Similarities and differences between normal-type and disease-type synapses are identified, and underlying synaptic phenotype groups are discovered and characterized. The results provide potential biologic insights into autism and schizophrenia that can serve as a starting point for further experimental analysis.
MIT Department
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Terms of Use
In Copyright - Educational Use Permitted
Copyright MIT
Persistent DSpace Link
