Measurements of ear-canal cross-sectional areas from live human ears with implications for wideband acoustic immittance measurements

Preliminary results of classifying otosclerosis and disarticulation using a convolutional neural network trained with simulated wideband acoustic immittance data

Current noninvasive methods of clinical practice often do not identify the causes of conductive hearing loss due to pathologic changes in the middle ear with sufficient certainty. Wideband acoustic immittance (WAI) measurement is noninvasive, inexpensive and objective. It is very sensitive to pathologic changes in the middle ear and therefore promising for diagnosis. However, evaluation of the data is difficult because of large interindividual variations. Machine learning methods like Convolutional neural networks (CNN) which might be able to deal with this overlaying pattern require a large amount of labeled measurement data for training and validation. This is difficult to provide given the low prevalence of many middle-ear pathologies. Therefore, this study proposes an approach in which the WAI training data of the CNN are simulated with a finite-element ear model and the Monte-Carlo method. With this approach, virtual populations of normal, otosclerotic, and disarticulated ears were generated, consistent with the averaged data of measured populations and well representing the qualitative characteristics of individuals. The CNN trained with the virtual data achieved for otosclerosis an AUC of 91.1 %, a sensitivity of 85.7 %, and a specificity of 85.2 %. For disarticulation, an AUC of 99.5 %, sensitivity of 100 %, and specificity of 93.1 % was achieved. Furthermore, it was estimated that specificity could potentially be increased to about 99 % in both pathological cases if stapes reflex threshold measurements were used to confirm the diagnosis. Thus, the procedures' performance is comparable to classifiers from other studies trained with real measurement data, and therefore the procedure offers great potential for the diagnosis of rare pathologies or early-stages pathologies. The clinical potential of these preliminary results remains to be evaluated on more measurement data and additional pathologies.

Reliability of an extended version of the 3m™ Eargage tool to assess earcanal size and assist earplug selection

OBJECTIVE

Evaluate the ability of an extended version of the 3 MTM Eargage to estimate the earcanal size and assess the likelihood that a particular earplug can fit an individual's earcanal, ultimately serving as a tool for selecting earplugs in the field.

DESIGN

Earcanal morphology, assessed through earcanal earmolds scans, is compared to earcanal size assessed with the extended eargage (EE) via box plots and Pearson linear correlations coefficients. Relations between attenuation measured on participants (for 6 different earplugs) and their earcanal size assessed with the EE are established via comparison tests.

STUDY SAMPLE

121 participants exposed to occupational noise (103 men, 18 women, mean age 47 years).

RESULTS

The earcanal size assessed with the EE allows for estimating the area of the earcanal's first bend cross-section (correlation coefficient   r = 0.533, p < 0.001). Extremely large earcanals (12.7% of earcanals in our sample) lead to significantly lower earplug attenuation (potentially inadequate) than smaller earcanals.

CONCLUSIONS

The EE is a simple and inexpensive tool easily deployable in the field to assist earplugs selection. When extended with sizes larger than the maximum size of the commercial tool, it allows for detecting individuals with extremely large earcanals who are most likely to be under-protected.

Classification, registration and segmentation of ear canal impressions using convolutional neural networks

Today, fitting bespoke hearing aids involves injecting silicone into patients' ears to produce ear canal molds. These are subsequently 3D scanned to create digital ear canal impressions. However, before digital impressions can be used they require a substantial amount of effort in manual 3D editing. In this article, we present computational methods to pre-process ear canal impressions. The aim is to create automation tools to assist the hearing aid design, manufacturing and fitting processes as well as normalizing anatomical data to assist the study of the outer ear canal's morphology. The methods include classifying the handedness of the impression into left and right ear types, orienting the geometries onto the same coordinate system sense, and removing extraneous artifacts introduced by the silicone mold. We investigate the use of convolutional neural networks for performing these semantic tasks and evaluate their accuracy using a dataset of 3000 ear canal impressions. The neural networks proved highly effective at performing these tasks with 95.8% adjusted accuracy in classification, 92.3% within 20° angular error in registration and 93.4% intersection over union in segmentation.

The influence of tympanic-membrane orientation on acoustic ear-canal quantities: A finite-element analysis

Assuming plane waves, ear-canal acoustic quantities, collectively known as wideband acoustic immittance (WAI), are frequently used in research and in the clinic to assess the conductive status of the middle ear. Secondary applications include compensating for the ear-canal acoustics when delivering stimuli to the ear and measuring otoacoustic emissions. However, the ear canal is inherently non-uniform and terminated at an oblique angle by the conical-shaped tympanic membrane (TM), thus potentially confounding the ability of WAI quantities in characterizing the middle-ear status. This paper studies the isolated possible confounding effects of TM orientation and shape on characterizing the middle ear using WAI in human ears. That is, the non-uniform geometry of the ear canal is not considered except for that resulting from the TM orientation and shape. This is achieved using finite-element models of uniform ear canals terminated by both lumped-element and finite-element middle-ear models. In addition, the effects on stimulation and reverse-transmission quantities are investigated, including the physical significance of quantities seeking to approximate the sound pressure at the TM. The results show a relatively small effect of the TM orientation on WAI quantities, except for a distinct delay above 10 kHz, further affecting some stimulation and reverse-transmission quantities.

Measurements of ear-canal geometry from high-resolution CT scans of human adult ears

Description of the ear canal's geometry is essential for describing peripheral sound flow, yet physical measurements of the canal's geometry are lacking and recent measurements suggest that older-adult-canal areas are systematically larger than previously assumed. Methods to measure ear-canal geometry from multi-planar reconstructions of high-resolution CT images were developed and applied to 66 ears from 47 subjects, ages 18-90 years. The canal's termination, central axis, entrance, and first bend were identified based on objective definitions, and the canal's cross-sectional area was measured along its canal's central axis in 1-2 mm increments. In general, left and right ears from a given subject were far more similar than measurements across subjects, where areas varied by factors of 2-3 at many locations. The canal areas varied systematically with age cohort at the first-bend location, where canal-based measurement probes likely sit; young adults (18-30 years) had an average area of 44mm2 whereas older adults (61-90 years) had a significantly larger average area of 69mm2. Across all subjects ages 18-90, measured means ± standard deviations included: canals termination area at the tympanic annulus 56±8mm2; area at the canal's first bend 53±18mm2; area at the canal's entrance 97±24mm2; and canal length 31.4±3.1mm2.

A reciprocity method for validating acoustic ear-probe source calibrations

Measurements of wideband acoustic immittance (WAI) rely on the calibration of an ear probe to obtain its acoustic source parameters. The clinical use of WAI and instruments offering the functionality are steadily growing, however, no international standard exists to ensure a certain reliability of the hardware and methods underlying such measurements. This paper describes a reciprocity method that can evaluate the accuracy of and identify errors in ear-probe source calibrations. By placing the ear probes of two calibrated WAI instruments face-to-face at opposite ends of a short waveguide, the source parameters of each ear probe can be measured using the opposite calibrated ear probe. The calibrated and measured source parameters of each ear probe can then be compared directly, and the influence of possible calibration errors on WAI measurements may be approximated. In various exemplary ear-probe calibrations presented here, the reciprocity method accurately identifies errors that would otherwise remain undetected and result in measurement errors in real ears. The method is likely unsuitable for routine calibration of WAI instruments but may be considered for conformance testing as part of a potential future WAI standard.