Refine
Document Type
- Conference proceeding (9)
- Journal article (7)
Is part of the Bibliography
- yes (16)
Institute
- Technik (16)
Publisher
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.
Background: Wideband acoustic immittance (WAI) and wideband tympanometry (WBT) are promising approaches to improve diagnosis accuracy in middle-ear diagnosis, though due to significant interindividual difference, their analysis and interpretation remains challenging. Recent approaches have come up, implementing machine learning (ML) or deep learning classifiers trained with measured WAI or WBT data for the classification of otitis media or otosclerosis. Also, first approaches have been made in identifying important regions from the WBT data, which the classifiers used for their decision-making.
Methods: Two classifiers, a convolutional neural network (CNN) and the ML algorithm extreme gradient boosting (XGB), are trained on artificial data obtained with a finite-element ear model providing the middle-ear measurements energy reflectance (ER), pressure reflectance phase, impedance amplitude and phase. The performance of both classifiers is evaluated by cross-validation on artificial test data and by classification of real measurement data from the literature using the metrics macro-recall and macro-F1 score. The feature contributions are quantified using the feature importance ‘gain’ for XGB and deep Taylor decomposition for CNN.
Results: In the cross-validation with artificial data, the macro-recall and macro-F1 scores are similar, namely 91.2% for XGB and 94.5% for CNN. For the classification with real measurement data the macro-recall and macro-F1-score were 81.8% and 38.2% (XGB) and 81.0% and 54.8% (CNN), respectively. The key features identified are ER between 600–1,000 Hz together with impedance phase between 600–1,000 Hz for XGB and ER up to 1,500 Hz for CNN.
Conclusions: We were able to show that the applied classifiers CNN and XGB trained with simulated data lead to a reasonably well performance on real data. We conclude that using simulation-based WAI data can be a successful strategy for classifier training and that XGB can be applied to WAI data. Furthermore, ML interpretability algorithms are useful to identify relevant key features for differential diagnosis and to increase confidence in classifier decisions. Further evaluation using more measured data, especially for pathological cases, is essential.
Machine learning algorithms and neural networks have recently been used for the classification of middle ear disorders using wideband acoustic immittance and wideband tympanometry data. This study applies the extreme gradient boosting (XGB) classifier, trained on simulated WAI data, to classify real measured data for normal, otosclerotic, and disarticulated ears. The achieved macro recall of 82 % is comparable to other approaches trained with real measurement data. The interpretability methods LIME and SHAP are used to quantify each feature’s contribution, both revealing energy reflectance between 600-800 Hz as a key feature for all classes. The key feature identified matches the differences that can be visually observed in the training and test data. However, the obtained feature contributions don’t provide enough distinguishable information to recognise incorrect or uncertain classifications.
Due to the large interindividual variances and the poor optical accessibility of the ear, the specificity of hearing diagnostics today is severely restricted to a certain clinical picture and quantitative assessment. Often only a yes or no decision is possible, which depends strongly on the subjective assessment of the ENT physician. A novel approach, in which objectively obtainable, non invasive audiometric measurements are evaluated using a numerical middle ear model, makes it possible to make the hidden middle ear properties visible and quantifiable. The central topic of this paper is a novel parameter identification algorithm that combines inverse fuzzy arithmetic with an artificial neural network in order to achieve a coherent diagnostic overall picture in the comparison of model and measurement. Its usage is shown at a pathological pattern called malleus fixation where the upper ligament of the malleus is pathologically stiffened.
This study describes a non-contact measuring and parameter identification procedure designed to evaluate inhomogeneous stiffness and damping characteristics of the annular ligament in the physiological amplitude and frequency range without the application of large static external forces that can cause unnatural displacements of the stapes. To verify the procedure, measurements were first conducted on a steel beam. Then, measurements on an individual human cadaveric temporal bone sample were performed. The estimated results support the inhomogeneous stiffness and damping distribution of the annular ligament and are in a good agreement with the multiphoton microscopy results which show that the posterior-inferior corner of the stapes footplate is the stiffest region of the annular ligament. This method can potentially help to establish a correlation between stiffness and damping characteristics of the annular ligament and inertia properties of the stapes and, thus, help to reduce the number of independent parameters in the model-based hearing diagnosis.
This study describes a non-contact measuring and system identification procedure for evaluating inhomogeneous stiffness and damping characteristics of the annular ligament in the physiological amplitude and frequency range without the application of large static external forces that can cause unnatural displacements of the stapes. To verify the procedure, measurements were first conducted on a steel beam. Then, measurements on an individual human cadaveric temporal bone sample were performed. The estimated results support the inhomogeneous stiffness and damping distribution of the annular ligament and are in a good agreement with the multiphoton microscopy results which show that the posterior-inferior corner of the stapes footplate is the stiffest region of the annular ligament.
Investigation of tympanic membrane influences on middle-ear impedance measurements and simulations
(2020)
This study simulates acoustic impedance measurements in the human ear canal and investigates error influences due to improperly accounted evanescence in the probe’s near field, cross-section area changes, curvature of the ear canal, and pressure inhomogeneities across the tympanic membrane, which arise mainly at frequencies above 10 kHz. Evanescence results from strongly damped modes of higher order, which can only be found in the near field of the sound source and are excited due to sharp cross-sectional changes as they occur at the transition from the probe loudspeaker to the ear canal. This means that different impedances are measured depending on the probe design. The influence of evanescence cannot be eliminated completely from measurements, however, it can be reduced by a probe design with larger distance between speaker and microphone. A completely different approach to account for the influence of evanescence is to evaluate impedance measurements with the help of a finite element model, which takes the precise arrangement of microphone and speaker in the measurement into account. The latter is shown in this study exemplary on impedance measurements at a tube terminated with a steel plate. Furthermore, the influences of shape changes of the tympanic membrane and ear canal curvature on impedance are investigated.
Perforations of the tympanic membrane (TM) can occur as a result of injury or inflammation of the middle ear. These perforations can lead to conductive hearing loss (HL), where in some cases the magnitude of HL exceeds that attributable to the observed TM perforation alone. We aim with this study to better understand the effects of location and size of TM perforations on the sound transmitting properties of the middle ear.
The middle ear transfer function (METF) of six human temporal bones (TB; freshly frozen specimen of body donors) were compared before and after perforation of the TM at different locations (anterior or posterior lower quadrant) and of different sizes (1mm, ¼ of the TM, ½ of the TM, and full ablation). The
METF were correlated with a Finite Element (FE) model of the middle ear, in which similar alterations were simulated.
The measured and simulated FE model METFs exhibited frequency and perforation size dependent amplitude losses at all locations and severities. In direct comparison, posterior TM perforations affected the transmission properties to a larger degree than perforations of the anterior quadrant. This could possibly be caused by an asymmetry of the TM, where the malleus-incus complex rotates and results in larger deflections in the posterior TM half than in the anterior TM half. The FE model of the TM with a sealed cavity suggest that small perforations result in a decrease of TM rigidity and thus to an increase in oscillation amplitude of the TM, mostly above 1 kHz.
The location and size of TM perforations influence the METF in a reproducible way. Correlating our data with the FE model could help to better understand the pathologic mechanisms of middle-ear diseases. If small TM perforations with uncharacteristically significant HL are observed in daily clinical practice, additional middle ear pathologies should be considered. Further investigations on the loss of TM pretension due to perforations may be informative.
Simulation models of the middle ear have rarely been used for diagnostic purposes due to their limited predictive ability with respect to pathologies. One big challenge is the large uncertainty and ambiguity in the choice of material parameters of the model.
Typically, the model parameters are determined by fitting simulation results to validation measurements. In a previous study, it was shown that fitting the model parameters of a finite-element model using the middle-ear transfer function and various other measurable output variables from normal ears alone is not sufficient to obtain a good predictive ability of the model on pathological middle-ear conditions. However, the inclusion of validation measurements on one pathological case resulted in a very good predictive ability also for other pathological cases. Although the found parameter set was plausible in all aspects, it was not yet possible to draw conclusions about the uniqueness and the accuracy or the uncertainty of the parameter set.
To answer these questions, statistical solution approaches are used in this study. Using the Monte Carlo method, a large number of plausible model data sets are generated that correctly represent the normal and pathological middle-ear characteristics in terms of various output variables like e.g., impedance, reflectance, umbo, and stapes transfer function. Subsequent principal component analyses (PCA) allow to draw conclusions about correlations, quantitative limits and statistical density of parameter values.
Furthermore, applying inverse PCA yields numerous plausible parameterizations of the middle-ear model, which can be used for data augmentation and training of a neural network which is capable of distinguishing between a normal middle ear and pathologies like otosclerosis, malleus fixation, and disarticulation based on objectively measured quantities like impedance, reflectance, and umbo velocity.
Development of a stochastic finite element model for use in the diagnosis of middle-ear pathologies
(2024)
The calibrated model accurately reproduces the mean and variance of middle-ear measurements like impedance, reflectance, stapes and umbo transfer function. Ligament and joint material parameters have a significant effect on the variance of these measurements, while variations in center of mass positions, for example, have less effect. The neural network trained on the simulated data shows promise for diagnostics, achieving 86-100% sensitivity and 85-93% specificity for detecting otosclerosis and disarticulation, which is similar to the performance of classifiers trained on measured immittance data.
