Krishna Mukunda

Mentored by IDIES member Kemar Green

Vertigo has a lifetime prevalence of around 7.4%, and is significantly difficult to diagnose. The presence or absence of torsional nystagmus can allow for quick diagnosis and prevent unnecessary medical testing. Our proposal aims to develop and explain a deep learning model for torsional identification from video oculography (VOG) recordings.

This approach could potentially reveal new clinical markers of torsion, improving the accessibility and accuracy of vestibular diagnostics. The model will differentiate between torsional nystagmus, static torsion, and non-torsion. Additionally, we will create a synthetic dataset using generative AI methods to develop an open dataset for broader research use.

KRISHNA MUKUNDA is a third-year Biomedical Engineering student from Cleveland, OH. His interests lie in integrating deep learning and neuroscience to enhance medical diagnostics. Post-graduation, he aspires to become a doctor, leveraging his engineering background to advance patient care.