Objective:Study on the value of quantitative measurement of key structures of inner ear by high resolution computer tomography(HRCT) in hearing evaluation of patients with congenital severe sensorineural hearing loss(SNHL). Methods:A total of 90 children with extremely severe SNHL diagnosed and treated in Beilun District People's Hospital from January 2018 to February 2021 were collected as the experimental group. In the same period, 90 children(180 ears) with normal inner ear structure and hearing were scanned because of head trauma and suspected temporal bone fracture. Logistic univariate and multivariate regression analysis were used to analyze the factors affecting the occurrence of extremely severe SNHL. Based on the results of multivariate analysis, a Nomogram prediction model was established. The model before and after internal correction was evaluated by the receiver working characteristic curve. Results:Inner ear malformation, SSCC bone island width, LSCC bone island width and cochlear height were independent risk factors for extremely severe SNHL. The results of Nomogram predictive model showed that cochlear height 34 points, LSCC bone island width 19 points, SSCC bone island width 22 points, inner ear malformation 37 points, the total score(112 points) corresponding to the incidence of extremely severe SNHL(0.3%). The actual C-index value of Nomogram prediction model is 0.858, the C-index of internal verification is 0.851, and the C-index of external verification is 0.847. The coincidence of the model is good. It is suggested that the model can effectively predict the risk factors of congenital extremely severe SNHL and has high prediction accuracy. Conclusion:The standardized measurement of SSCC bone island width, LSCC bone island width and cochlear height by HRCT is of great value in the diagnosis of microinner ear malformation in children with extremely severe SNHL.