Abstract:Objective:To explore the predictive value of video-head impulse test (vHIT) in the degree of recovery of high-frequency function of the semicircular canal in patients with vestibular neuritis (VN). Methods A total of 147 VN patients who visited our department from September 2019~March 2024 were selected as the study subjects, including 75 males and 72 females, aged 18~75 years, with an average age of (36.45±8.21) years, and were divided into good recovery group (n=81) and poor recovery group (n=66) according to the degree of recovery of high-frequency function of the semicircular canal. The clinical data of the two groups and the vHIT test indicators were compared. The receiver operating characteristic curve (ROC) was used to evaluate the predictive value of each index on the degree of recovery of high-frequency function of the semicircular canal. Multivariate logistic regression analysis was performed to analyze the independent influencing factors of high-frequency functional recovery of semicircular canals, and a logistic regression model was constructed.Evaluate the model"s discrimination using the ROC curve, examine the model"s predictive accuracy with the calibration curve, and assess the model"s clinical value through decision curve analysis. Use the enhanced Bootstrap method for internal validation of the model and evaluate its performance during internal validation. Results The proportion of cervical spondylosis, migraine, stroke and sleep disorder in the poor recovery group was higher than that in the recovery group, and the vestibular-ocular reflex (VOR) gain value in the poor recovery group was lower than that in the recovery group, and the positive rates of saccade wave wave, vHIT abnormality rate and asymmetric gain value were higher than those in the recovery group, and the difference was statistically significant(P<0.05). The results of ROC curve evaluation showed that vHIT indexes had certain guiding value in predicting the recovery of high-frequency function of the semicircular canal after treatment (P<0.05, AUC>0.70), among which the accuracy of VOR gain value was high (P<0.05, AUC>0.80), and the optimal cut-off value was used as the boundary for parallel diagnosis, and the results showed that the combined judgment of each index had the highest accuracy in judging poor recovery (P<0.001, AUC>0.90). Multivariate Logistic regression analysis showed that migraine, stroke, VOR gain value, and VOR gain asymmetric ratio were independent predictors of the recovery of high-frequency function of the semicircular canal. Conclusions This study analyzed clinical characteristics and vHIT indicators, finding that migraine, stroke, VOR gain values, and gain asymmetry ratio are independent factors predicting the recovery of high-frequency semicircular canal function. Moreover, the combined model based on these indicators can effectively predict recovery status, providing important references for clinical intervention.
