Facebook Twitter VKontakte
Sign in
  • Home
  • Human Eye Project
    • The Pupil
    • Pupil Decentration-Multiformaties
    • Pupil Deformations
    • Pupil Color and Dimensions
    • Pupil Anisocoria
    • Pupil Miosis
    • Pupil Mydriasis
    • Pupil Reflexes
    • The Collarette
    • The Iris
  • PupilMetrics App
    • PupilMetrics Android
    • PupilMetrics Windows
    • PupilMetrics Mac OS
  • Documentation
  • Pricing
  • Learn More
    • CNRI Research
    • History
    • About
    • Eye Photo Tips & Tricks
    • IRB Status
    • Pupil Abstract Blog
    • Privacy Policy
    • Terms of Service
Sign in
Welcome!Log into your account
Forgot your password?
Password recovery
Recover your password
Search
Sign in
Welcome! Log into your account
Forgot your password? Get help
Password recovery
Recover your password
A password will be e-mailed to you.
Clinical Neuro-Optic Reseach Initiative
  • Home
  • Human Eye Project
    • The Pupil
    • Pupil Decentration-Multiformaties
    • Pupil Deformations
    • Pupil Color and Dimensions
    • Pupil Anisocoria
    • Pupil Miosis
    • Pupil Mydriasis
    • Pupil Reflexes
    • The Collarette
    • The Iris
  • PupilMetrics App
    • PupilMetrics Android
    • PupilMetrics Windows
    • PupilMetrics Mac OS
  • Documentation
  • Pricing
  • Learn More
    • CNRI Research
    • History
    • About
    • Eye Photo Tips & Tricks
    • IRB Status
    • Pupil Abstract Blog
    • Privacy Policy
    • Terms of Service

Getting Started

5
  • 1.1 System Requirements
  • 1.2 Installation
  • 1.3 Licensing
  • 1.4 First Launch
  • 1.5 Desktop Window & Keyboard Shortcuts

Capturing Eye Images

8
  • 2. Capturing Eye Images
  • 2.1 Camera Source Selection
  • 2.2 Quality-Gated Camera Mode – Android App
  • 2.3 Manual Camera Mode – Android
  • 2.4 USB / UVC Iriscope (Dino-Lite)
  • 2.5 PLR Video Mode – Android
  • 2.6 Import from Gallery
  • 2.7 Tips for a Good Capture

Reading the Analysis Results

10
  • 3. Reading the Analysis Results
  • 3.1 The Iris Zone Map
  • 3.2 PI Ratio (Pupil–Iris Ratio)
  • 3.3 Zone Findings — Flattenings (FLAT) and Protrusions (PROT)
  • 3.4 ANW Assessment (Collarette / Autonomic Nerve Wreath)
  • 3.5 Decentration (Pupil Position)
  • 3.6 Ellipseness (Pupil Shape)
  • 3.7 Anisocoria (Pupil Size Difference)
  • 3.8 Confidence Scores & Hybrid Fusion
  • 3.9 Scan History

Patient Management

4
  • 4. Patient Management & Exports
  • 4.1 PDF Report
  • 4.2 Plain-Text & JSON Export
  • 4.3 Sharing & Filing

Natural Medicine Therapy Panels

8
  • 5. Natural Medicine Therapy Panels
  • 5.1 Enabling the Therapy Modules
  • 5.2 How Zone Findings Drive the Therapy Panels
  • 5.3 Herbal Recommendations Panel
  • 5.4 Nutrition Recommendations Panel
  • 5.5 Chiropractic Correlations Panel
  • 5.6 TCM Correlations Panel
  • 5.7 Reading Therapy Panels Together

Constitutional Iridology

7
  • 6. Constitutional Iridology
  • 6.1 Background & Theoretical Basis
  • 6.2 The 34 Constitutional Types
  • 6.3 Selecting a Constitutional Type
  • 6.4 Constitutional Panel in Analysis Results
  • 6.5 Constitutional Section in the PDF Report
  • 6.6 Clinical Guidance & Limitations

Exporting PDF Reports

2
  • 7. Exporting PDF Reports
  • 7.1 Regenerating a PDF

Settings & Customization

5
  • 8. Settings & Customization
  • 8.1 Languages
  • 8.2 Zone Overlay & Observer Notes
  • 8.3 ML Comparison Panel
  • 8.4 About & Support

Clinical & Legal Disclaimers

2
  • 9. Clinical & Legal Disclaimers
  • 9.1 Data Privacy
View Categories
  • Home
  • PupilMetrics Documentation
  • Reading the Analysis Results
  • 3.8 Confidence Scores & Hybrid Fusion

3.8 Confidence Scores & Hybrid Fusion

1 min read

PupilMetrics runs **two independent analysis pipelines** on every image and then fuses their outputs into a single confidence score.

**Classical CV (pixel-based)**

The classical computer vision pipeline uses circle detection, radial sampling, and boundary-point analysis on the full-resolution image. It produces pixel-accurate iris and pupil boundaries.

**ML Model (ONNX)**

The machine learning model (`cnri_model.onnx`) is a neural network trained on iris images, resized to a normalized 224×224 crop centered on the detected iris. It outputs four regression values: PI ratio, decentration, ellipseness, and decentration angle.

**Hybrid Confidence Formula**

The four components are weighted and combined:

| Component | Weight | What it measures |

|———–|——–|—————–|

| Capture quality | 20% | Image sharpness, brightness, contrast from the quality gate |

| Classical CV confidence | 35% | Circle detection score from the Hough-like iris finder |

| ML plausibility | 20% | Whether the ML outputs fall within anatomically reasonable ranges |

| Cross-model agreement | 25% | How closely the two pipelines agree on PI ratio (80%), ellipseness (10%), and decentration (10%) |

The fused confidence is displayed as a percentage and maps to the familiar grade:

| Fused confidence | Grade |

|—————–|——-|

| > 75% | **A** |

| 60–75% | **B** |

| 45–60% | **C** |

| < 45% | **D** |

**Safety caps**

If classical CV confidence falls below 25%, or if capture quality falls below 30%, the fused score is capped at 40% or 50% respectively, regardless of other components. This ensures a poor underlying image always produces a conservative grade.

**When the two pipelines disagree**

When the classical and ML results differ significantly on PI ratio (> 10% tolerance), the agreement component reduces the hybrid score. The individual classical and ML values are still shown in the results for reference when “Show ML Comparison” is enabled in settings.

Updated on March 24, 2026

What are your Feelings

  • Happy
  • Normal
  • Sad

Share This Article :

  • Facebook
  • X
  • LinkedIn
  • Pinterest
3.7 Anisocoria (Pupil Size Difference)3.9 Scan History

The Clinical Neuro-Optic Research Initiative (CNRI) advances pupil-based neurodiagnostics by preserving historical insights, developing modern analytic tools, and researching links between ocular microstructures and systemic health. Our mission is to validate and expand neuro-optic biomarkers for breakthroughs in early detection, monitoring, and non-invasive assessment of autonomic and neurological function.

Twitter