Roshni Parulekar-Martins
Using Convolutional Neural Networks to Learn About the Brain's Neural Circuitry
2025–2026
Electrical Engineering and Computer Science
- Health and Life Sciences
You, Sixian
How do we think and form memories? Understanding this requires mapping neuronal firing and synaptic dynamics in the hippocampus, the brain’s center for learning and memory. Voltage imaging, which records neural activity via fluorescent genetically encoded voltage indicators (GEVIs), enables functional readout of these processes. While prior work has focused on structural reconstructions or broad behavioral dynamics, few have used voltage imaging to infer synaptic strengths. This project develops a pipeline to reconstruct hippocampal circuits from voltage imaging movies and track how repeated stimulation reshapes connection weights, offering new insights into the neural basis of learning and memory.
I’m excited to conduct research on the “final frontier,” understanding how our brains work. SuperUROP will be an amazing opportunity to learn hands-on experimental and computational techniques, deepen my understanding of brain function, and contribute to a project that could one day impact BCIs, AI, and treatments for memory disorders.
