Related Experiment Video
Updated: Aug 6, 2026
07:24
Using Eye-tracking to Assess the Relative Importance of Visual and Vestibular Input to Subcortical Motion Processing in the Roll Plane
Published on: August 22, 2025
Posterior Canal and Atypical Benign Paroxysmal Positional Vertigo: Development of a Predictive Model for Clinical
Kathrine M Jakobsen1, Line M Nielsen1, Clara Bender1
1Department of Health Science and Technology, Aalborg University, Aalborg, DNK.
Cureus
|July 23, 2026
Summary
A new data-driven model accurately distinguishes typical and atypical Benign Paroxysmal Positional Vertigo (BPPV) using eye movement data. This tool aids in precise diagnosis for dizziness, improving patient outcomes.
Area of Science:
- Neurology
- Medical Diagnostics
- Data Science in Medicine
Background:
- Benign Paroxysmal Positional Vertigo (BPPV) is a common cause of dizziness, but clinical diagnosis is often challenging.
- Difficulty in discerning eye movement direction and intensity leads to diagnostic uncertainty and suboptimal treatment for BPPV.
Purpose of the Study:
- To develop and evaluate a data-driven model for differentiating typical and atypical variants of BPPV.
- To improve diagnostic accuracy and clinical decision-making in BPPV cases.
Main Methods:
- Retrospective analysis of data from 26 BPPV patients and 2 controls.
- Extraction and processing of eye and head movement data using VisualEyes™ and MATLAB.
- Logistic regression model utilizing latency and peak slow-phase velocity (SPV) of torsional nystagmus, evaluated with ROC analysis and cross-validation.
Main Results:
- The model achieved an Area Under the ROC Curve (AUC) of 0.936, indicating excellent discrimination between typical and atypical BPPV.
- Cross-validation yielded high performance metrics: 88.0% accuracy, 78.2% sensitivity, and 93.9% specificity.
- The model demonstrated 100% agreement with expert clinical assessments on unseen patient data.
Conclusions:
- The developed data-driven model exhibits high diagnostic accuracy for differentiating BPPV variants.
- The model shows significant promise as a clinical decision support tool for BPPV diagnosis.
