Middle- and inner-ear surgery is a vital treatment option in hearing loss, infections, and tumors of the lateral skull base. Segmentation of otologic structures from computed tomography (CT) has many potential applications for improving surgical planning but can be an arduous and time-consuming task. We propose an end-to-end solution for the automated segmentation of temporal bone CT using convolutional neural networks (CNN). Using 150 manually segmented CT scans, a comparison of 3 CNN models (AH-Net, U-Net, ResNet) was conducted to compare Dice coefficient, Hausdorff distance, and speed of segmentation of the inner ear, ossicles, facial nerve and sigmoid sinus. Using AH-Net, the Dice coefficient was 0.91 for the inner ear; 0.85 for the ossicles; 0.75 for the facial nerve; and 0.86 for the sigmoid sinus. The average Hausdorff distance was 0.25, 0.21, 0.24 and 0.45 mm, respectively. Blinded experts assessed the accuracy of both techniques, and there was no statistical difference between the ratings for the two methods (p = 0.93). Objective and subjective assessment confirm good correlation between automated segmentation of otologic structures and manual segmentation performed by a specialist. This end-to-end automated segmentation pipeline can help to advance the systematic application of augmented reality, simulation, and automation in otologic procedures.

Objective: The objective of this study is to build upon previous work validating a tablet-based software to measure cochlear duct length (CDL). Here, we do so by greatly expanding the number of cochleae (n = 166) analyzed, and examined whether computed tomography (CT) slice thickness influences reliability of CDL measurements. Study Design: Retrospective chart review study. Setting: Tertiary referral center. Patients: Eighty-three adult cochlear implant recipients were included in the study. Both cochleae were measured for each patient (n = 166). Interventions: Three raters analyzed the scans of 166 cochleae at 2 different time points. Each rater individually identified anatomical landmarks that delineated the basal turn diameter and width. These coordinates were applied to the elliptic approximation method (ECA) to estimate CDL. The effect of CT scan slice thickness on the measurements was explored. Main Outcome Measures: The primary outcome measure is the strength of the inter- and intra-rater reliability. Results: The mean CDL measured was 32.84 ± 2.03 mm, with a range of 29.03 to 38.07 mm. We observed no significant relationship between slice thickness and CDL measurement (F1,164 = 3.04; p = 0.08). The mean absolute difference in CDL estimations between raters was 1.76 ± 1.24 mm and within raters was 0.263 ± 0.200 mm. The intra-class correlation coefficient (ICC) between raters was 0.54 and ranged from 0.63 to 0.83 within raters. Conclusions: This software produces reliable measurements of CDL between and within raters, regardless of CT scan thickness.

Objective To assess whether health literacy is associated with: 1) degree of hearing loss at initial presentation for audiogram and 2) hearing aid adoption for hearing aid candidates. Methods We identified 1376 patients who underwent audiometric testing and completed a brief health literacy questionnaire at our institution. The association between health literacy and degree of hearing loss at initial presentation was examined using linear regression, adjusted for age, gender, marital status, education level, race, language, employment status, and insurance coverage. The association between health literacy and hearing aid adoption was examined in the subset of patients identified as hearing aid candidates using logistic regression, adjusted for demographic factors and insurance coverage. Results Patients with inadequate health literacy were more likely to present with more severe hearing loss (adjusted mean pure‐tone average [PTA] difference, 5.38 dB, 95% confidence interval [CI] 2.75 to 8.01). For hearing aid candidates (n = 472 [41.6%]), health literacy was not associated with hearing aid adoption rate (odds ratio [OR] 0.85, 95% CI 0.40 to 1.76). Hearing aid coverage through Medicaid (OR 2.22, 95% CI 1.13 to 4.37), and moderate (OR 2.70, 95% CI 1.58 to 4.69) or moderate–severe (OR 2.23, 95% CI 1.19 to 4.16) hearing loss were associated with hearing aid adoption. Conclusions In our population, patients with low health literacy are more likely to present with higher degrees of hearing loss, but no less likely to obtain hearing aids compared with patients with adequate health literacy. Hearing loss severity and hearing aid coverage by insurance appear to be the main drivers of hearing aid adoption. Level of Evidence 3 Laryngoscope, 131:E1688–E1694, 2021

The H(+) hypothesis of lateral feedback inhibition in the outer retina predicts that depolarizing agents should increase H(+) release from horizontal cells. To test this hypothesis, self-referencing H(+) -selective microelectrodes were used to measure extracellular H(+) fluxes from isolated goldfish horizontal cells. We found a more complex pattern of cellular responses than previously observed from horizontal cells of other species examined using this technique. One class of cells had an initial standing signal indicative of high extracellular H(+) adjacent to the cell membrane; challenge with glutamate, kainate or high extracellular potassium induced an extracellular alkalinization. This alkalinization was reduced by the calcium channel blockers nifedipine and cobalt. A second class of cells displayed spontaneous oscillations in extracellular H(+) that were abolished by cobalt, nifedipine and low extracellular calcium. A strong correlation between changes in intracellular calcium and extracellular proton flux was detected in experiments simultaneously monitoring intracellular calcium and extracellular H(+) . A third set of cells was characterized by a standing extracellular alkalinization which was turned into an acidic signal by cobalt. In this last set of cells, addition of glutamate or high extracellular potassium did not significantly alter the proton signal. Taken together, the response characteristics of all three sets of neurons are most parsimoniously explained by activation of a plasma membrane Ca(2+) ATPase pump, with an extracellular alkalinization resulting from exchange of intracellular calcium for extracellular H(+) . These findings argue strongly against the hypothesis that H(+) release from horizontal cells mediates lateral inhibition in the outer retina.

Conclusion: Our results suggest that accounting for 4 kHz oVEMP amplitude can improve detection of SCD compared to the binary presence of n10 response. The 4 kHz oVEMP amplitude cutoff that maximizes sensitivity and specificity for our cohort is 15 uV. Our results also suggest that 4 kHz oVEMP amplitudes align better with symptomatic SCD cases compared to cases in which there is radiographic SCD but no characteristic symptoms.