- Role
- Finalist
- Region
- York
- Pronouns
- she/her/hers
Aashika Reddy
ea adipisicing consectetur ex in sit enim duis consequat exercitation officia sit fugiat eiusmod ex cupidatat in laborum dolore amet magna sunt culpa ea pariatur consectetur enim ex consequat laborum laborum irure exercitation culpa non amet
Project
HARMONY: A Machine Learning-Based Adaptive Music Neurostimulation Device for Pediatric Epilepsy
Project #5306
Pediatric epilepsy presents a critical medical challenge, affecting millions of children suffering from unpredictable, intractable seizures with limited efficacy of existing therapies. HARMONY addresses this gap by fusing machine learning with personalized music therapy for real‑time seizure detection and intervention. Using the CHB‑MIT EEG dataset (22 pediatric subjects, ages 1.5–22), a Random Forest model enhanced with SMOTE was developed, achieving 90% recall and 98% accuracy in live seizure prediction. Upon detection, the system instantly plays original, patient-preferred music compositions across multiple genres. Each composition is meticulously designed to replicate the therapeutic properties of Mozart’s Sonata K.448 with >95% similarity, harnessing the biologically validated "Mozart effect" to prevent and/or reduce seizure intensity. HARMONY presents a groundbreaking, non-invasive algorithm that offers a cost-effective, AI-driven music therapy solution for pediatric epilepsy management, promising significant real‑world impact and an enhanced quality of life for young patients.
- Challenge
- Disease and Illness
- Category
- Senior
- Type
- Innovation