Validation of a portable video head impulse test using an iPod touch and an oral stabilization three-dimensional (3D) - printed mount: a method-comparison study

Full abstract

Read the full abstract

Background Although video head impulse testing (vHIT) is widely used to assess semicircular canal function, access to quantitative vestibular testing remains limited. Smartphone-based vHIT systems with goggle-based fixation have been described, but larger validation studies and improved fixation stability are needed. We evaluated a portable vHIT application using an iPod touch with an oral stabilization three-dimensional (3D)-printed mount to determine whether it detects laterality in unilateral vestibular hypofunction and is noninferior to a medical-grade device for a patient-level laterality endpoint. Methods Thirteen patients with unilateral vestibular hypofunction and 14 healthy volunteers underwent testing with the portable and medical-grade vHIT systems on the same day. The primary endpoint was the patient-level laterality difference, defined as the difference between unaffected and affected vestibulo-ocular reflex (VOR) gain. Noninferiority was assessed using the two-sided 95% confidence interval (CI) for the mean between-device difference with a prespecified margin of 0.15. Agreement was evaluated using Bland–Altman analysis. Secondarily, absolute VOR gain agreement between devices was evaluated. False-positive classifications in healthy volunteers were assessed at the ear and subject levels using a gain threshold of <0.78. Results Both systems showed higher VOR gains on the unaffected side (paired t -test: medical device, p = 2.44 × 10 −7 ; portable vHIT, p = 1.02 × 10 −8 ). Mean laterality difference was 0.518 ± 0.180 for the medical-grade device and 0.453 ± 0.118 for the portable system. The between-device difference was −0.065 (95% CI: −0.149 to 0.018), supporting noninferiority. Bland–Altman analysis demonstrated a bias of −0.065, with 95% limits of agreement from −0.337 to 0.207. For absolute VOR gain, average between-device bias was small, but individual-ear agreement remained variable. In healthy volunteers, abnormal classifications occurred in 0 of 28 ears and 0 of 14 subjects with the medical-grade device, and in 2 of 28 ears (7.1%) and 2 of 14 participants (14.3%) with the portable system. Conclusion The portable vHIT system detected laterality in unilateral vestibular hypofunction and met noninferiority criteria relative to a medical-grade device. Larger studies are warranted to refine quality control and define individual-ear absolute VOR gain agreement.