Humans are able to hear in a variety of complicated acoustic environments. This feat begins in the peripheral auditory system, where the cochlea collects and transmits thousands of individual bits of sound data to the brain. Here, we introduce GSP Cochlea: a graph signal processing-based framework to investigate and visualize sound encoding. We show that a cochlea graph with a mesh topology provides a mechanism of denoising, efficient information transfer, and modular processing. We demonstrate an application to assess hearing loss as more than just a decibel loss at particular frequencies of sound but also a significant change to the cochlea graph architecture. GSP Cochlea is a generalized approach that provides new insight into the higher-level functional activity of the inner ear.

ImportanceProgress in vestibular implantation offers hope for patients with bilateral vestibulopathy. However, surgically opening the semicircular canals risks breaching the membranous labyrinth, which may induce sensorineural hearing loss. A robotic drill sensing force and torque might prevent membranous labyrinth rupture.Primary objectiveTo assess the feasibility of force- and torque-based automatic cessation in a hand-guided robotic drill for fenestrating the bony semicircular canals without rupturing the membranous labyrinth.Secondary objectiveTo fit an electrode dummy through the fenestrations.DesignFeasibility study using human cadaveric temporal bones.SettingLaboratory.ParticipantsTen formalin-fixed human temporal bones.InterventionAfter performing a cortical mastoidectomy and skeletonizing the semicircular canals, a hand-guided robotic drill was used to drill 2 fenestrations in each semicircular canal. A silicone electrode dummy was inserted through each fenestration.Main outcome measuresProportion of fenestrations with intact membranous labyrinth, as evaluated with a surgical microscope. Proportion of fenestrations allowing electrode insertion without additional manipulation.ResultsA total of 60 fenestrations were made in 30 semicircular canals from 10 temporal bones. Technical issues related to drill bit fixation occurred in 6 fenestrations. The remaining 54 fenestrations were all made without visible damage to the membranous labyrinth. In 81% of these fenestrations (44/54), the electrode could be advanced without requiring additional manipulation. The technical issue was related to improper alignment of the drill bit, leading to incorrect force and torque sensing.ConclusionsForce- and torque-based automatic cessation in a hand-guided robotic drill is feasible for fenestrating the bony semicircular canals without rupturing the membranous labyrinth. However, improved burr fixation is required for consistent and reliable performance.RelevanceThe investigated approach holds potential to improve safety and precision in semicircular canal surgery, such as vestibular implantation. This may expand treatment options for patients with residual inner ear function.

With the global expansion of industrial sectors, noise has emerged as a significant harmful physical agent in workplaces. Besides auditory effects, high-level noise exposure can cause non-auditory consequences, especially at high sound intensities. These effects may include metabolic disturbances such as Metabolic Syndrome (MetS). This study aimed to examine the relationship between noise-induced hearing loss (NIHL) at different sound frequencies and the occurrence of MetS and its defining factors. This cross-sectional descriptive-analytical study was conducted on 1,142 male employees in a petrochemical industry. Fasting blood glucose, triglycerides, and HDL cholesterol were measured from serum samples, and pure-tone audiometry was performed to assess hearing thresholds at 500, 1000, 2000, 3000, 4000, 6000, and 8000 Hz. MetS was defined based on the NCEP-ATP III criteria. Data were analyzed using logistic regression models, with MetS as the dependent variable and NIHL as the predictor, adjusting for age, work experience, BMI, and blood pressure. Findings showed the prevalence of MetS and NIHL was 6.92% and 22.24%, respectively. mean hearing threshold across the tested frequencies significantly higher in individuals with MetS than in those without (p < 0.05), and this difference was more pronounced at higher frequencies. Statistical analysis indicated that NIHL was associated with higher odds of MetS (OR = 1.35, 95% CI: 1.27-1.44; p < 0.05), although the strength of this association decreased with increasing frequency. The greatest NIHL effect was seen in raising triglyceride (TG) levels, while its effect on lowering HDL levels was not statistically significant (p > 0.05). Among demographic variables, higher BMI and education level were associated with higher odds of MetS. Overall, the results show that frequency-specific patterns of NIHL were associated with a higher likelihood of MetS. Therefore, controlling noise and addressing its non-auditory effects is essential in industrial settings.

This study aimed to evaluate risk factors for neonatal hearing loss among newborns referred to tertiary audiology centers and to investigate their association with confirmed hearing outcomes. A total of 1,869 newborns referred to tertiary audiology centers after failed newborn hearing screening or due to established risk factors between January 2021 and January 2023 were retrospectively evaluated using the OtoAccess database. The analysis was restricted to referred newborns rather than the entire birth cohort. During 2021-2022, a total of 113,665 births were recorded in Şanlıurfa, of which 1,869 newborns were referred for diagnostic evaluation at tertiary audiology centers. Hearing loss was detected in 8.4% of infants. Among newborns with at least one documented risk factor, 48 (3.5%) met the clinical criteria for cochlear implantation. Family history of hearing loss and syndromic disorders were significantly associated with neonatal hearing loss (p < 0.05). No statistically significant differences were observed between infants with and without hearing loss in terms of mechanical ventilation, ototoxic drug exposure, hyperbilirubinemia, neurodegenerative disease, TORCH infections, cerebral pathology, or maternal diseases. Nearly half of affected infants had severe-to-profound hearing loss, highlighting the importance of early cochlear implant evaluation in referred neonatal populations. Family history of hearing loss and syndromic disorders were identified as significant factors associated with neonatal hearing loss in this referral-based cohort. The high prevalence of consanguineous marriage in the region may be associated with the burden of familial hearing loss. Additionally, the occurrence of hearing loss in infants without identifiable risk factors underscores the importance of universal newborn hearing screening and timely referral to specialized audiology centers.

Ear disease contributes significantly to global hearing loss, with recurrent otitis media being a primary preventable cause in children, impacting development. Artificial intelligence (AI) offers promise for early diagnosis via otoscopic image analysis, but dataset biases and inconsistencies limit model generalizability and reliability. This retrospective study systematically evaluated three public otoscopic image datasets (Chile; Ohio, USA; Türkiye) using quantitative and qualitative methods. Two counterfactual experiments were performed: (1) obscuring clinically relevant features to assess model reliance on non-clinical artifacts, and (2) evaluating the impact of hue, saturation, and value on diagnostic outcomes. Quantitative analysis revealed significant biases in the Chile and Ohio, USA datasets. Counterfactual Experiment I found high internal performance (AUC > 0.90) but poor external generalization, because of dataset-specific artifacts. The Türkiye dataset had fewer biases, with AUC decreasing from 0.86 to 0.65 as masking increased, suggesting higher reliance on clinically meaningful features. Counterfactual Experiment II identified common artifacts in the Chile and Ohio, USA datasets. A logistic regression model trained on clinically irrelevant features from the Chile dataset achieved high internal (AUC = 0.89) and external (Ohio, USA: AUC = 0.87) performance. Qualitative analysis identified redundancy in all the datasets and stylistic biases in the Ohio, USA dataset that correlated with clinical outcomes. In summary, dataset biases significantly compromise reliability and generalizability of AI-based otoscopic diagnostic models. Addressing these biases through standardized imaging protocols, diverse dataset inclusion, and improved labeling methods is crucial for developing robust AI solutions, improving high-quality healthcare access, and enhancing diagnostic accuracy.