[Full-frequency hearing curve analysis in patients with sensorineural hearing loss]

Objective: To combine the conventional audiometric curves and extended high frequency audiometric curves of patients with sensorineural deafness to form a full-frequency audiometric curve and to perform typing, so as to comprehensively understand the hearing status of patients with sensorineural deafness. Methods: This study was a cross-sectional study. The study subjects included 249 patients with sensorineural hearing loss who visited the Otolaryngology-Head and Neck Surgery outpatient clinic of the PLA General Hospital between July 2019 and December 2020. Among them, 146 were male and 103 were female, aged 11 to 80 years. The cases included 123 with mild hearing loss, 70 with moderate hearing loss, 32 with moderately severe hearing loss, 17 with severe hearing loss, 6 with profound hearing loss, and 1 with total deafness. According to the national standard GB/T16403-1996, conventional pure-tone audiometry (125-8 000 Hz) was performed on the 249 patients with sensorineural hearing loss to obtain their conventional-frequency hearing curves, which were then classified. Extended high-frequency pure-tone threshold testing (9 000-20 000 Hz) was conducted using extended high-frequency headphones, specifically including eight frequencies: 9 000, 10 000, 11 200, 12 500, 14 000, 16 000, 18 000, and 20 000 Hz. Ultimately, the full-frequency hearing curves (125-20 000 Hz) of each patient were obtained. The K-means clustering analysis method was used to classify the hearing curves based on their characteristics, and the results of the K-means clustering analysis were partially adjusted through manual screening. Results: The conventional hearing curves of all 249 patients were consistent with sensorineural hearing loss. The detection rates for extended high frequencies (9 000, 10 000, 11 200, 12 500, 14 000, 16 000, 18 000, and 20 000 Hz) were 96.79% (241/249), 94.38% (235/249), 87.95% (219/249), 78.31% (195/249), 65.46% (163/249), 22.09% (55/249), 10.84% (27/249), and 0.80% (2/249), respectively. The conventional-frequency hearing curves of the patients could be classified into the following types: low-frequency descending type (50/249, 20.08%), conventional-frequency steep descending type (78/249, 31.33%), conventional-frequency gradual descending type (58/249, 23.29%), conventional-frequency flat type (25/249, 10.04%), conventional-frequency 4 000 Hz notch type (30/249, 12.05%), and other types (8/249, 3.21%). By incorporating extended high-frequency hearing data, the full-frequency hearing curves of 235 patients were further classified into the following types based on different characteristics: full-frequency hill type (32/235, 13.62%), full-frequency ascending type (28/235, 11.91%), full-frequency 8 000 Hz notch type (14/235, 5.96%), full-frequency steep descending type (82/235, 34.89%), full-frequency gradual descending type (34/235, 14.47%), full-frequency shoulder-raising type (7/235, 2.98%), full-frequency shoulder-dropping type (25/235, 10.64%), full-frequency flat type (8/235, 3.40%), and other full-frequency types (5/235, 2.13%). Conclusions: Compared to the classification based on conventional-frequency hearing curves, the full-frequency hearing curves of patients with sensorineural hearing loss provide a more comprehensive representation of their overall hearing status. Patients with the same conventional-frequency hearing curve classification may exhibit different full-frequency hearing curve types, suggesting potential differences in the location and extent of pathological damage within their auditory systems.