# Wearable Sensor–Based Gait Analysis in Benign Paroxysmal Positional Vertigo: Quantitative Assessment of Residual Dizziness Using the φ-Bonacci Framework

## TL;DR

This study uses wearable sensors and a new framework called φ-bonacci to detect gait issues and residual dizziness in patients with BPPV before and after treatment.

## Contribution

The study introduces the φ-bonacci index framework to objectively quantify gait impairments and residual dizziness in BPPV patients.

## Key findings

- BPPV patients showed higher self-similarity gait metrics than controls before treatment.
- Post-treatment, gait metrics normalized, but residual dizziness was linked to elevated stability metrics under eyes-closed conditions.
- Visual deprivation increased gait instability metrics across all groups.

## Abstract

Background: Benign Paroxysmal Positional Vertigo (BPPV) is the most common vestibular disorder. Although canalith repositioning procedures (CRPs) typically resolve positional vertigo, several patients still report imbalance or residual dizziness, which remain difficult to quantify with standard tests. Wearable inertial sensors now allow high-resolution, objective gait analysis and may detect subtle vestibular-related impairments. Objectives: This study evaluates the clinical usefulness of sensor-based gait metrics, enhanced by the newly developed φ-bonacci index framework to quantify gait changes and residual dizziness in BPPV before and after CRPs. Methods: Fifteen BPPV patients (BPPV-P) and fifteen age-matched controls completed walking tests under eyes-open (EO) and eyes-closed (EC) conditions using wearable inertial measurement units (IMU). φ-bonacci index components—self-similarity (A1), swing symmetry (A2), and double-support consistency (A4)—were calculated to assess gait harmonicity, symmetry and stability. Results: Before treatment, BPPV-P exhibited significantly higher A1 values than healthy controls (p = 0.038 EO; p = 0.011 EC), indicating impaired gait harmonicity. After CRPs, A1 values normalized to control levels, suggesting restored gait self-similarity. Under visual deprivation, both A1 and A4 showed pronounced increases across all groups, reflecting the contribution of vision to balance control. Among post-treatment patients, those reporting residual dizziness demonstrated persistently elevated A4 values—particularly under EC conditions—indicating incomplete sensory reweighting despite clinical recovery. Conclusions: Wearable sensor–derived φ-bonacci metrics offer sensitive, objective markers of gait abnormalities and residual dizziness in BPPV, supporting their use as digital biomarkers for diagnosis, rehabilitation, and follow-up.