We're BrainScores. AI-powered diagnostics that reveal surgical targets in epilepsy patients

We're BrainScores. AI-powered diagnostics that reveal surgical targets in epilepsy patients

Millions live with seizures no medication can control. Yet only 1% receive curative surgery

20 million

people globally suffer from pharmaco-resistant epilepsy (~100k in Canada)

20 years

lost by patients cycling through medications before surgical referral

Only 1%

of eligible patients undergo surgery each year due to diagnostic uncertainty

The barrier between invisible and undetectable isn't medical—it's technological. We built the technology to eliminate that barrier

Approach

BrainScores was built specifically for epilepsy surgery. Generic neuroimaging platforms serve multiple conditions but lack the depth required for focal cortical dysplasia detection

Domain expertise and annotated dataset

Years of epilepsy imaging research have defined the patterns of focal cortical dysplasia and subtle structural abnormalities that standard protocols miss. Over 4,000 epilepsy cases with confirmed surgical outcomes comprise the largest annotated training dataset in the field.

Domain expertise and annotated dataset

Years of epilepsy imaging research have defined the patterns of focal cortical dysplasia and subtle structural abnormalities that standard protocols miss. Over 4,000 epilepsy cases with confirmed surgical outcomes comprise the largest annotated training dataset in the field.

Multi-modal architecture and integration

Multiple MRI sequences (T1, FLAIR, diffusion, and functional imaging) processed together capture the specific signatures of epileptic tissue. These sequences reveal tissue patterns that remain invisible to single-modality analysis.

Multi-modal architecture and integration

Multiple MRI sequences (T1, FLAIR, diffusion, and functional imaging) processed together capture the specific signatures of epileptic tissue. These sequences reveal tissue patterns that remain invisible to single-modality analysis.

Domain expertise and annotated dataset

Years of epilepsy imaging research have defined the patterns of focal cortical dysplasia and subtle structural abnormalities that standard protocols miss. Over 4,000 epilepsy cases with confirmed surgical outcomes comprise the largest annotated training dataset in the field.

Multi-modal architecture and integration

Multiple MRI sequences (T1, FLAIR, diffusion, and functional imaging) processed together capture the specific signatures of epileptic tissue. These sequences reveal tissue patterns that remain invisible to single-modality analysis.

Automated detection with expert validation

AI analysis identifies lesions expert radiologists cannot see, highlighting surgical targets with precision. International epilepsy specialists review every diagnostic output to ensure clinical accuracy. AI assists clinical judgment but never replaces it.

Automated detection with expert validation

AI analysis identifies lesions expert radiologists cannot see, highlighting surgical targets with precision. International epilepsy specialists review every diagnostic output to ensure clinical accuracy. AI assists clinical judgment but never replaces it.

Open science foundation and peer review

Published, peer-reviewed algorithms rather than proprietary black boxes ensure transparency. Core components remain open to scientific scrutiny. Six software toolboxes have achieved 40,000+ downloads and 1,500+ citations across the research community.

Open science foundation and peer review

Published, peer-reviewed algorithms rather than proprietary black boxes ensure transparency. Core components remain open to scientific scrutiny. Six software toolboxes have achieved 40,000+ downloads and 1,500+ citations across the research community.

Automated detection with expert validation

AI analysis identifies lesions expert radiologists cannot see, highlighting surgical targets with precision. International epilepsy specialists review every diagnostic output to ensure clinical accuracy. AI assists clinical judgment but never replaces it.

Open science foundation and peer review

Published, peer-reviewed algorithms rather than proprietary black boxes ensure transparency. Core components remain open to scientific scrutiny. Six software toolboxes have achieved 40,000+ downloads and 1,500+ citations across the research community.

Clinical Validation

Patients who spent years MRI-negative now receive curative surgery

BrainScores was built exclusively for epilepsy surgery. Generic neuroimaging platforms serve multiple conditions but lack the depth required for focal cortical dysplasia detection.

Example #1

Focal cortical dysplasia

Standard MRI:

Negative

BrainScores analysis:

Lesion detected in right frontal cortex

Surgical outcome:

Seizure-free at 12 months

MRI-negative
Automated deviation mapping
AI diagnostics
Curative surgery

Example #1

Focal cortical dysplasia

Standard MRI:

Negative

BrainScores analysis:

Lesion detected in right frontal cortex

Surgical outcome:

Seizure-free at 12 months

MRI-negative
Automated deviation mapping
AI diagnostics
Curative surgery

Example #2

Temporal lobe epilepsy

Standard MRI:

Negative

BrainScores analysis:

Hippocampal abnormality identified

Surgical outcome:

Seizure-free at 18 months

MRI-negative
Automated deviation mapping
AI diagnostics
Curative surgery

Example #2

Temporal lobe epilepsy

Standard MRI:

Negative

BrainScores analysis:

Hippocampal abnormality identified

Surgical outcome:

Seizure-free at 18 months

MRI-negative
Automated deviation mapping
AI diagnostics
Curative surgery

Platform

Four integrated capabilities for epilepsy surgical planning. From raw MRI data to surgical navigation in a single integrated system

01

Processing. Multi-modal MRI standardization and quality control form the open-source foundation available for research use

02

Analysis. Whole-brain morphometric mapping reveals deviation patterns across cortical and subcortical structures

03

Diagnostics. AI-powered lesion identification combined with expert radiologist validation delivers comprehensive case reports

04

Surgical integration. Direct data export to neuronavigation systems from Medtronic, ROSA, Brainlab, and Zimmer Biomet

Team

Built by the researchers who defined modern epilepsy imaging. Founded by recognized leaders in epilepsy imaging research

Boris Bernhardt, PhD

Co-founder, Project Lead

Associate Professor of Neurology and Neurosurgery, Canada Research Chair, Montreal Neurological Institute

Boris Bernhardt, PhD

Co-founder, Project Lead

Associate Professor of Neurology and Neurosurgery, Canada Research Chair, Montreal Neurological Institute

Raul Cruces, MD PhD

Co-founder, Neuroinformatics Lead

Research Associate, Montreal Neurological Institute

Raul Cruces, MD PhD

Co-founder, Neuroinformatics Lead

Research Associate, Montreal Neurological Institute

Boris Bernhardt, PhD

Co-founder, Project Lead

Associate Professor of Neurology and Neurosurgery, Canada Research Chair, Montreal Neurological Institute

Raul Cruces, MD PhD

Co-founder, Neuroinformatics Lead

Research Associate, Montreal Neurological Institute

Jordan De Kraker, PhD

Co-founder, AI Lead

Postdoctoral Researcher, Montreal Neurological Institute

Jordan De Kraker, PhD

Co-founder, AI Lead

Postdoctoral Researcher, Montreal Neurological Institute

Philippe Benjamin-Duranceau

Co-founder, Operations lead

Philippe Benjamin-Duranceau

Co-founder, Operations lead

Jordan De Kraker, PhD

Co-founder, AI Lead

Postdoctoral Researcher, Montreal Neurological Institute

Philippe Benjamin-Duranceau

Co-founder, Operations lead

Enning Yang

Research Software developer

Montreal Neurological Institute

Enning Yang

Research Software developer

Montreal Neurological Institute

Ian Goodall-Halliwell

Software developer

PhD student, Montreal Neurological Institute

Ian Goodall-Halliwell

Software developer

PhD student, Montreal Neurological Institute

Ella Sahlas

Data Analyst

MD-PhD student, Montreal Neurological Institute

Ella Sahlas

Data Analyst

MD-PhD student, Montreal Neurological Institute

Judy Chen

Data Analyst

MD-PhD student, Montreal Neurological Institute

Judy Chen

Data Analyst

MD-PhD student, Montreal Neurological Institute

Enning Yang

Research Software developer

Montreal Neurological Institute

Ian Goodall-Halliwell

Software developer

PhD student, Montreal Neurological Institute

Ella Sahlas

Data Analyst

MD-PhD student, Montreal Neurological Institute

Judy Chen

Data Analyst

MD-PhD student, Montreal Neurological Institute

Research
leadership

Research leadership

Decades of leading international task forces and ILAE consortia in epilepsy imaging.

4k+
cases

4k+ cases

Database built through collaborations with 15+ leading epilepsy centers worldwide.

40k+
downloads

Six published software toolboxes with 1,500+ citations across the research community.

40k+ downloads

Six published software toolboxes with 1,500+ citations across the research community.

Integration

We advance epilepsy diagnostics through clinical partnerships, research collaboration, and open science

For clinical partners

Schedule a call to discuss your institution's needs and review case demonstrations relevant to your patient population.

For research collaborators

We partner with epilepsy centers on validation studies and novel diagnostic approaches. Active projects underway in Montreal, Toronto, New York, and Boston.

For talent

We're building technology that changes lives. If that matters to you, we should talk.

For talent

We're building technology that changes lives. If that matters to you, we should talk.

@ BrainScores AI LLM

@ BrainScores AI LLC

2025–Future

Multimodal Imaging and Connectome Analysis Lab
Montreal Neurological Institute and Hospital
3801 University Street, Montreal, QC, H3A 2B4, Canada

+1 514 398 3579

Create a free website with Framer, the website builder loved by startups, designers and agencies.