Dhruv S. 2026 | BASIS Independent Silicon Valley
- Project Title: Discovering EEG Biomarkers Through Applications of EEG and Machine Learning for Genetic Epilepsies (SYNGAP1)
- BASIS Independent Advisor: Allendoerfer
- Internship Location: Stanford School of Medicine
- Onsite Mentor: Christopher Lee-Messer
My name is Dhruv Sharma, and I’m a 12th grader at BASIS Independent Silicon Valley. My project focuses on discovering novel EEG biomarkers in genetic epilepsies through applications of machine learning and EEG analysis. This project is currently being researched in the Dr. Lee-Messer Lab at Stanford University, where I am an intern specifically working on this study.
The research will begin with library and internet research, during which I will analyze and understand scientific papers to build my prior knowledge. These papers include signal-processing techniques, programs for detecting specific EEG biomarkers such as HFOs, and lectures detailing particular cases of epileptic seizures and how to diagnose them from EEG. From there, I will conduct observations and experiments using machine learning processes and tools such as PyTorch, Morgoth, SpikeNet2 (EEG machine-learning models), and PyHFO to identify novel EEG biomarkers in epileptic mice with the SYNGAP1 genetic mutation.
After extracting features, we will compare healthy mice with SYNGAP1 mice to identify specific novel biomarkers, and then apply the drug nortriptyline to the SYNGAP1 mice to determine whether it affects or reduces epileptic seizures relative to initial hypotheses. As of now, I have been working with PyHFO, a computational framework for detecting high-frequency oscillations (HFOs) in EEG signals, examining how these patterns may identify seizure-generating networks and predict epileptogenic treatment (the start of seizure) responses. Alongside collaborators in the Stanford Knowles Lab, I will then also studying how spike-wave discharges progress over time in SYNGAP1+/− mice and whether drugs like nortriptyline can actually restore normal hippocampal function by reducing pathological HFO activity, (this is one of our hypotheses)
In the coming months, the final product of this research is that we expect to complete this EEG biomarker–discovery project and research paper, and I will be presenting our findings at BASIS and, with my research team, at the American Clinical Neurophysiology Society (ACNS) conference, contributing to the broader scientific effort to identify new EEG biomarkers for epileptic encephalopathies. Furthermore, this research will ignite and deepened my interest in computational neuroscience, a field I'm very passionate in pursuing and learning more into.
