top of page

How did we design NEUROFUSE?

Design

Design 1:
Pain as a Biomarker

  • Focused on correlation with disease

  • Descriptive, retrospective

  • ❌ No targeting, no regulation

Design 2:
Simulated Metabolic Models

  • Introduced machine learning

  • Relied on simulated mitochondrial dysfunction

  • ❌ Models learned simulation artifacts, not physiology

Design 3:
Imaging + ML

  • Used MRSI features

  • Treated brain regions independently

  • ❌ No modeling of network propagation

Final Design:
NEUROFUSE

  • Combines metabolism + connectivity

  • Identifies drivers vs bystanders

  • Enables closed-loop adaptation

Required Breakthroughs

  • Higher-resolution MRSI

  • More precise connectome mapping

  • Improved interpretability between signatures and biology

  • Longitudinal stability validation

NEUROFUSE

Research Project

by Stuyvesant High School Research Team

© 2035 by Neurofuse. Powered and secured by Wix 

bottom of page