top of page

Precise & InterpretableBrain Tumor Diagnosis

Open-source AI to revolutionize neuro-ethics and energy-efficiency in early brain tumor diagnosis

heatmap.jpg

Featured at

image_edited.png

Revolutionizing Life: Frontiers of Biotechnology Conference 2025

Published in

image.png

1st Place at

image.png

Harvard VISION Global Health Leadership Conference Pitch Competition

Grid

What is Neuroflux?

Neuroflux analyzes brain scans, classifies the tumor type, and produces a heatmap highlighting tumor damage.

MRI_brain_scan.gif

Neuroflux Version 2.0

Neuroflux is now compatible with CT scans, accepting the scans as inputs and outputting a heat map and tumor probability with a 97.69% accuracy. This reflects an ongoing effort to bring brain tumor diagnosis to a wider population.

Screenshot 2025-09-16 at 4.40.54 PM.png

Visualize

Neuroflux in Action

Slide to reveal how
Neuroflux's heat maps visually pinpoint
tumor regions

pexels-mart-production-7089331_edited.jpg

Why Neuroflux Matters

Restorative Yoga

Neuroflux focuses on glioblastoma multiforme, one of the most aggressive brain tumors. Due to vague early symptoms and diagnostic delays, GBM often goes undetected until advanced stages and significantly reduces survival rates.

Grid

Better for Doctors, Better for Patients

Neuroflux accurately identifies brain tumor location and spread, giving doctors clear visual insights for faster decisions. By supporting earlier diagnosis, Neuroflux also helps patients start treatment sooner and improves their chances of recovery.

pexels-mart-production-7089020_edited_edited.jpg
WhatsApp Image 2025-02-19 at 7.46_edited.jpg

SDG 3

Good Health and Well-Being

01

Designed to detect GBM earlier and more accurately, enabling faster diagnosis and improving patient survival outcomes

02

Developed as a lightweight, open-source software to ensure access to AI diagnostics for low-resource hospitals without costly computational infrastructure

03

Incorporates interpretable heatmaps to empower clinicians in verifying diagnostic decisions, boosting clinical transparency

04

Enables radiologists to precisely pinpoint tumor boundaries and plan safer, more effective treatments for patients

Accelerating SDG 3 Targets

Target 3.4: Reduce Premature Mortality

Reducing mortality rates through AI-powered early tumor detection

Target 3.8: Achieve Universal Health Coverage

Expanding access to AI diagnostic tools in underserved regions

What Clinicians &
Researchers Are Saying

"This is an outstanding research project conducted by a group of high school students who are very well-read in their field of computational biology. They demonstrate a college-level and above understanding of how machine learning models are built and tested for human diseases, particularly for neuroscience."
bottom of page