How Nurodot Works
Combining advanced neuroimaging, cognitive assessment, and AI to revolutionize Alzheimer's detection and monitoring.
Advanced Alzheimer's Risk Assessment
Nurodot provides a comprehensive platform for early detection and monitoring of Alzheimer's disease by analyzing MRI scans and cognitive test results.
The Alzheimer's Challenge
Alzheimer's disease affects over 6 million Americans, with numbers expected to triple by 2050. Early detection is critical, but traditional methods face significant challenges:
- Manual MRI analysis is time-consuming and subjective
- Cognitive tests are often performed inconsistently
- Integration of biomarkers and cognitive data is complex
- Results can take weeks to months to process
Nurodot solves these problems with an integrated platform that combines advanced imaging analysis with standardized cognitive testing—delivering results in minutes instead of months.
Our Approach
We provide a comprehensive assessment by combining three essential elements:
- Advanced MRI Analysis: Using FastSurfer, a deep learning-based tool, to extract clinically validated biomarkers.
- Standardized Cognitive Testing: Implementing digital versions of MMSE, CDR, and ADAS-Cog with consistent scoring.
- AI-Powered Integration: Generating comprehensive reports that synthesize imaging and cognitive data into actionable insights.
FastSurfer: Advanced Neuroimaging Analysis
What is FastSurfer?
FastSurfer is a deep learning-based neuroimaging analysis pipeline developed by researchers at Harvard Medical School and the German Center for Neurodegenerative Diseases. It delivers FreeSurfer-quality brain segmentation and analysis in under 5 minutes—compared to 8-10 hours with traditional methods.
Why We Use FastSurfer
- Speed: Analyzes full MRI scans in minutes instead of hours
- Accuracy: Maintains clinical-grade precision with deep learning
- Consistency: Eliminates subjective interpretation between different radiologists
- Reliability: Validated against gold-standard FreeSurfer measurements
- Scalability: Enables population-level screening with GPU acceleration
FastSurfer segments the brain into distinct regions for biomarker extraction
FastSurfer delivers similar accuracy to FreeSurfer in a fraction of the processing time
The Technical Process
MRI Upload
Users upload standard NIFTI format (.nii or .nii.gz) MRI brain scans to our secure platform. These are typically T1-weighted 3D sequences—the same scans already acquired in standard neuroimaging protocols.
Deep Learning Segmentation
FastSurfer's convolutional neural networks (CNNs) rapidly segment the brain into distinct anatomical regions, including cortical and subcortical structures. This process takes 1-2 minutes on our GPU-accelerated infrastructure.
Biomarker Extraction
From the segmented brain regions, we calculate volumes, thicknesses, and other quantitative metrics that serve as biomarkers for neurodegeneration. These include hippocampal volume, ventricular expansion, and cortical thickness.
Clinical Contextualization
The extracted biomarkers are compared against normative databases, including the Alzheimer's Disease Neuroimaging Initiative (ADNI), to provide age-matched context for the measurements.
Clinical Biomarkers
We extract and analyze multiple neuroimaging biomarkers that have been validated in clinical research for Alzheimer's detection.
Hippocampal Volume
The hippocampus is critical for memory formation and is one of the earliest brain regions affected by Alzheimer's disease. We measure both left and right hippocampal volumes and calculate total hippocampal volume.
Normal range: 6000-8000 mm³ (total)
MCI range: 5000-6000 mm³, AD range: <5000 mm³
Ventricular Enlargement
As brain tissue atrophies in Alzheimer's, the fluid-filled ventricles expand. We measure ventricular volume and calculate the Evans Index (ratio of ventricle width to skull width).
Normal Evans Index: <0.3
Abnormal Evans Index: >0.3
Cortical Thickness
Thinning of the cerebral cortex occurs in Alzheimer's, particularly in regions involved in memory and cognition. We measure thickness across 34 cortical regions.
Normal mean thickness: >2.5mm
AD range: <2.4mm in key regions
Entorhinal Cortex
The entorhinal cortex is often the first region to show atrophy in Alzheimer's. We measure both its volume and thickness as early indicators.
Normal thickness: >3.2mm
MCI range: 2.8-3.2mm, AD range: <2.8mm
Asymmetry Index
Alzheimer's disease often affects brain regions asymmetrically. We calculate indices that measure this asymmetry, particularly in hippocampal and temporal lobe structures.
Normal range: <0.03
Abnormal range: >0.05
Whole Brain Volume
Overall brain volume decreases in Alzheimer's disease. We measure total brain volume and compare it against age-matched controls.
Normal range: Age-dependent
AD range: 2-5% below age-matched controls
ADNI Reference Values
We compare patient measurements against the Alzheimer's Disease Neuroimaging Initiative (ADNI) database, which contains standardized values for healthy controls, MCI, and Alzheimer's patients.
| Biomarker | Healthy Controls | Mild Cognitive Impairment | Alzheimer's Disease |
|---|---|---|---|
| Total Hippocampal Volume | 6998 ± 765 mm³ | 6038 ± 993 mm³ | 4970 ± 959 mm³ |
| Entorhinal Thickness | 3.45 ± 0.28 mm | 3.04 ± 0.43 mm | 2.68 ± 0.51 mm |
| Mean Cortical Thickness | 2.73 ± 0.13 mm | 2.58 ± 0.18 mm | 2.42 ± 0.20 mm |
| Ventricular Volume | 35,400 ± 19,800 mm³ | 42,900 ± 22,200 mm³ | 52,600 ± 24,600 mm³ |
| Evans Index | 0.28 ± 0.04 | 0.31 ± 0.04 | 0.35 ± 0.05 |
Standardized Cognitive Assessment
Why Cognitive Testing Matters
Cognitive testing provides crucial information that complements neuroimaging. While MRI shows structural changes, cognitive tests reveal functional deficits in memory, language, and executive function.
We implement digital versions of three gold-standard assessments:
- MMSE (Mini-Mental State Examination) - A 30-point test assessing orientation, attention, memory, language, and visual-spatial skills.
- CDR (Clinical Dementia Rating) - Evaluates cognitive and functional performance in six domains: memory, orientation, judgment, community affairs, home & hobbies, and personal care.
- ADAS-Cog (Alzheimer's Disease Assessment Scale-Cognitive) - A detailed assessment specifically designed to evaluate the severity of cognitive symptoms in Alzheimer's.
Understanding Binary Scoring (0 or 1)
Our platform uses binary scoring for cognitive test items, where:
- 1 point = Correct or expected response
- 0 points = Incorrect or impaired response
This binary approach has several advantages:
- Simplifies test administration in digital environments
- Reduces scoring ambiguity and subjective interpretation
- Enables automated calculation of total scores
- Mirrors the underlying scoring methods of standardized tests
- Facilitates machine learning analysis of cognitive patterns
MMSE Scoring
The MMSE contains 30 items across multiple cognitive domains. Each correct answer scores 1 point, with a maximum score of 30.
Interpretation:
24-30: Normal cognition
19-23: Mild cognitive impairment
10-18: Moderate impairment
≤9: Severe impairment
CDR Scoring
The CDR evaluates impairment severity in six domains. Our platform collects binary responses to questions in each domain, then calculates the standard CDR score (0-3).
Interpretation:
0: Normal
0.5: Very mild dementia
1: Mild dementia
2: Moderate dementia
3: Severe dementia
ADAS-Cog Scoring
The ADAS-Cog assesses multiple cognitive domains with a focus on memory, language, and praxis. Higher scores indicate greater impairment.
Interpretation:
0-10: Normal range
11-21: Mild impairment
22-35: Moderate impairment
>35: Severe impairment
Distribution of cognitive test scores across diagnostic categories
The Power of Combined Assessment
When cognitive testing is combined with neuroimaging biomarkers, diagnostic accuracy improves significantly. Studies show that the combination of these two approaches can improve early detection of Alzheimer's by up to 40% compared to either method alone.
Our platform captures both structural changes (via MRI) and functional deficits (via cognitive testing), providing a comprehensive assessment that supports more accurate diagnosis and better monitoring of disease progression.
AI-Powered Risk Analysis
Comprehensive Risk Assessment
Nurodot's AI analysis engine combines neuroimaging biomarkers and cognitive test scores to generate a detailed risk assessment for Alzheimer's disease. Our approach goes beyond simple classification to provide:
- Disease stage classification (Normal, MCI, Alzheimer's)
- Risk score quantification (0-100 scale)
- Neurodegeneration severity metrics
- Cognitive decline assessment
- 5-year progression risk estimation
- Comparison to ADNI reference data
Risk Analysis Methodology
Our risk analysis algorithm incorporates weighted contributions from multiple biomarkers and cognitive scores:
Neurodegeneration Score
Calculated from weighted biomarkers:
- Hippocampal volume (40%)
- Cortical thickness (25%)
- Entorhinal measurements (25%)
- Ventricular expansion (10%)
Cognitive Decline Score
Derived from test scores with these weights:
- MMSE (40%)
- CDR (40%)
- ADAS-Cog (20%)
Overall Risk Score
Combined from:
- Neurodegeneration (60%)
- Cognitive decline (40%)
Adjusted for age and other demographic factors
Breakdown of risk factors and their contribution to overall Alzheimer's risk assessment
GPT-Powered Clinical Summaries
Our platform uses GPT-4 (a large language model) to generate natural language clinical summaries that integrate all quantitative findings. This provides several benefits:
- Translation of complex biomarkers into understandable insights
- Contextual interpretation based on patient demographics and clinical history
- Clear explanation of risk levels and contributing factors
- Suggested next steps and follow-up recommendations
- Format that resembles traditional radiologist reports for easy integration into clinical workflows
Sample GPT-Generated Report
Alzheimer's Disease Assessment Report
Executive Summary:
The patient presents with neuroimaging biomarkers and cognitive performance consistent with Mild Cognitive Impairment (MCI). Hippocampal volumes are reduced compared to age-matched healthy controls, with some cortical thinning and mild ventricular enlargement. Current overall disease risk score is 58.4/100 with a five-year progression risk to Alzheimer's disease of 52.6%.
Neuroimaging Findings:
• Hippocampal Volumes: Left (2953 mm³) and Right (3085 mm³) show moderate atrophy
• Total Hippocampal Volume: 6038 mm³ (below age-adjusted norms)
• Asymmetry Index: 0.043 - Within expected parameters
• Evans Index: 0.31 - Mildly elevated, suggesting ventricular enlargement
• Mean Cortical Thickness: 2.58 mm - Shows modest thinning in AD-vulnerable regions
Recommendations:
1. Consider cognitive enhancement therapy based on clinical judgment
2. Comprehensive management of vascular risk factors
3. Structured cognitive training and physical exercise program
4. Follow-up neuropsychological testing in 6 months
Clinical Applications
Nurodot's platform supports multiple use cases across the Alzheimer's disease care continuum.
Early Detection
Identify subtle changes in brain structure and cognitive function before symptoms become pronounced. Early detection enables intervention at the MCI stage, when treatments are more likely to be effective.
Disease Monitoring
Track disease progression over time with consistent, quantitative measurements. This enables better treatment adjustment and helps patients and families prepare for future care needs.
Clinical Trial Screening
Identify suitable candidates for Alzheimer's clinical trials based on specific biomarker profiles and cognitive patterns. This improves trial recruitment efficiency and effectiveness.
Treatment Response Assessment
Evaluate the effectiveness of interventions by measuring changes in biomarkers and cognitive scores over time, providing objective evidence of treatment impact.
Benefits for Clinicians
- Reduced time spent on manual MRI analysis
- Consistent, objective measurements across patients
- Evidence-based risk assessment
- Clear, concise reports for clinical decision-making
- Longitudinal tracking of patient progression
Benefits for Patients
- Earlier detection of cognitive decline
- More accurate diagnosis
- Opportunity for earlier intervention
- Better understanding of disease progression
- Improved long-term care planning
Data Security & Privacy
At Nurodot, we understand the sensitivity of medical data. Our platform is built with security and privacy as foundational principles:
End-to-End Encryption
All data in transit and at rest is encrypted using industry-standard protocols, protecting sensitive information from unauthorized access.
Secure Cloud Infrastructure
Our platform runs on secure, isolated cloud infrastructure with regular security audits and penetration testing.
De-identified Analysis
MRI scans and cognitive data are processed using de-identified techniques, minimizing exposure of personal information.
