CN-122004849-A - Method and system for predicting hearing threshold by combining broadband middle ear impedance and otoacoustic emission

CN122004849ACN 122004849 ACN122004849 ACN 122004849ACN-122004849-A

Abstract

The invention discloses a method and a system for predicting hearing threshold by combining broadband middle ear impedance and otoacoustic emission, wherein the method comprises the steps of determining a detection mode and initializing detection parameters, sequentially executing ear canal sealing and environment noise self-detection, broadband middle ear impedance detection and otoacoustic emission detection, respectively extracting related characteristics of broadband middle ear impedance and related characteristics of otoacoustic emission, fusing the two types of characteristics to construct a combined characteristic, inputting a hearing prediction model, outputting a hearing threshold predicted value or a hearing condition evaluation result and generating a detection report, adopting a modularized framework of a core control and acquisition unit and an air pressure regulating unit, multiplexing core hardware to realize one-stop detection of middle ear and inner ear functions, and improving the accuracy of hearing threshold prediction, realizing subjective cooperation of a subject in the detection process by jointly modeling middle ear conduction characteristics and inner ear active functions and comprehensively covering key links of auditory paths.

Inventors

GONG QIN
Zhong Luliang
FAN YIBO

Assignees

清华大学
无锡清耳话声科技有限公司
浙江清华长三角研究院

Dates

Publication Date: 20260512
Application Date: 20260403

Claims (10)

1. A method of threshold prediction for wideband middle ear impedance in combination with otoacoustic emissions, comprising: determining a detection mode according to user selection or system preset rules, and initializing detection parameters corresponding to the detection mode; after finishing the steps of selecting the detection mode and initializing parameters, performing auditory canal sealing detection and environmental noise self-detection; After the preset sealing and noise conditions are met, wideband middle ear impedance detection is performed to obtain middle ear acoustic response changing with frequency and pressure, and wideband middle ear impedance related features are extracted from the middle ear acoustic response; After the wideband middle ear impedance detection is completed, performing otoacoustic emission detection to acquire otoacoustic emission signals and extracting otoacoustic emission related features therefrom; Performing fusion processing on the broadband middle ear impedance related characteristic and the otoacoustic emission related characteristic to construct a combined characteristic comprising middle ear conduction characteristic information and inner ear active function information; Inputting the joint characteristics into a hearing prediction model to output a hearing threshold predicted value or a hearing condition evaluation result; And generating and outputting a detection report according to the hearing threshold predicted value or the hearing condition evaluation result.
2. A method of threshold prediction in combination with wideband middle ear impedance and otoacoustic emissions as claimed in claim 1 wherein the selection of the detection mode includes the steps of: According to the selected detection mode, the system automatically loads preset parameters corresponding to the detection mode, wherein the detection mode comprises a hearing threshold prediction mode, a hearing condition screening mode and a middle ear conduction abnormality assessment mode; the preset parameters comprise general data quality control parameters, wideband middle ear impedance detection parameters and otoacoustic emission detection parameters, and various parameters support adjustment.
3. A method of threshold prediction in combination with wideband middle ear impedance and otoacoustic emissions as claimed in claim 1 wherein performing ear canal seal detection and environmental noise self-test comprises the steps of: the air pressure adjusting unit is used for carrying out tightness verification on the auditory meatus, collecting auditory meatus pressure data and analyzing pressure stability; collecting and evaluating background noise in the auditory canal, and judging whether a preset noise threshold condition is met; And when the sealing verification does not reach the standard or the background noise does not meet the preset condition, generating prompt information, and when the condition is met, continuing to execute the subsequent detection step.
4. A method of threshold prediction in combination with otoacoustic emissions for wideband middle ear impedance as claimed in claim 1 wherein said performing wideband middle ear impedance detection comprises the steps of: according to a preset air pressure scanning range and a preset stepping, sequentially performing pressure stabilizing regulation and control on each air pressure point; after each air pressure point is stable, outputting a broadband detection signal, and synchronously collecting a reflected acoustic signal; Calculating wideband middle ear impedance-related features based on the collected reflected acoustic signals, the wideband middle ear impedance-related features including wideband absorption, wideband reflection, admittance, or amplitude and phase features of the impedance; and carrying out data quality evaluation on the collected reflected acoustic signals, and recording measurement stability and reliability indexes.
5. A method of threshold prediction in combination with wideband middle ear impedance and otoacoustic emissions as claimed in claim 1 wherein said otoacoustic emissions detection comprises the steps of: Generating an otoacoustic emission induction stimulation signal, outputting the otoacoustic emission induction stimulation signal to the auditory canal of the subject, and synchronously collecting the returned otoacoustic emission signal; preprocessing the collected otoacoustic emission signals, and extracting otoacoustic emission related characteristics, wherein the otoacoustic emission related characteristics comprise an otoacoustic emission amplitude spectrum, a signal-to-noise ratio spectrum and fine structural characteristics; And performing data quality control on the extracted otoacoustic emission related features, recording signal-to-noise ratio and repeated consistency indexes, and marking low-quality data.
6. A method of threshold prediction for wideband middle ear impedance in combination with otoacoustic emissions as claimed in claim 1 wherein said combined signature construction comprises the steps of: Performing frequency axis alignment processing on the broadband middle ear impedance related features and the otoacoustic emission related features to enable the two types of features to be represented under the same or mappable frequency sampling points; Interpolation or resampling is carried out on the aligned characteristic data; Normalizing the processed characteristic data; Removing marked abnormal points or unreliable data, and complementing the missing information; and constructing a joint feature matrix or tensor for fusing middle ear conduction characteristic information and inner ear active function information.
7. A method of threshold prediction in combination with wideband middle ear impedance and otoacoustic emissions as claimed in claim 1 wherein said hearing condition assessment comprises the steps of: Inputting the combined features into a double-branch fusion hearing prediction model; independent high-order feature extraction is carried out on the relevant features of the middle ear impedance and the relevant features of the otoacoustic emission respectively through the double branches of the hearing prediction model; weighting and fusing the extracted two branch characteristics through a fusion layer, and carrying out joint reasoning by combining a multi-task learning structure; and outputting a corresponding evaluation result according to the detection mode, wherein the evaluation result comprises a hearing threshold predicted value, a hearing loss probability, a hearing grading result or a middle ear conduction abnormal risk score of each target frequency point.
8. A method of threshold prediction in combination with wideband middle ear impedance and otoacoustic emissions as claimed in claim 1 wherein said resultant output comprises the steps of: Visually displaying the broadband middle ear impedance related characteristic, the otoacoustic emission related characteristic and the evaluation result in the detection process; Integrating detection parameters, the broadband middle ear impedance related features, the otoacoustic emission related features, the measurement stability and reliability indexes, the signal to noise ratio, the repetition consistency indexes and the evaluation results to generate a standardized detection report; The detection report is stored to a local or remote server and supports export or printing.
9. A broad-band ear impedance-otoacoustic emission combined threshold prediction system for implementing the threshold prediction method according to any one of the preceding claims 1 to 8, characterized by comprising a core control and acquisition unit and an air pressure adjustment unit; the core control and acquisition unit consists of a computer, a universal acquisition sound card and an integrated acoustic probe; the computer is internally provided with a signal analysis program, a characteristic fusion program and a hearing prediction model; the air pressure adjusting unit is connected with the core control and acquisition unit and is used for adjusting the air pressure of the auditory canal in the detection process and providing a controllable air pressure environment for wideband middle ear impedance detection; The system is configured to perform wideband middle ear impedance detection through the cooperation of the air pressure adjusting unit and the core control and acquisition unit in a wideband middle ear impedance detection mode to extract wideband middle ear impedance related features; The system carries out fusion processing on the broadband middle ear impedance related characteristic and the otoacoustic emission related characteristic through a characteristic fusion program built in the computer, constructs a joint characteristic, analyzes the joint characteristic through the hearing prediction model, and outputs a hearing threshold predicted value or a hearing condition evaluation result.
10. An electronic device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the steps of the hearing threshold prediction method of any one of claims 1 to 8.

Description

Method and system for predicting hearing threshold by combining broadband middle ear impedance and otoacoustic emission Technical Field The invention relates to the technical field of objective detection and hearing evaluation of auditory systems, in particular to a detection system and a detection method for combining a broadband middle ear impedance detection technology with an otoacoustic emission detection technology and carrying out hearing threshold prediction and hearing condition evaluation based on joint characteristics. Background The pure tone threshold test is currently the most widely used hearing test method clinically, which determines the threshold level by subjective feedback of subjects on pure tone stimuli of different frequencies and intensities. However, the method is highly dependent on subjective coordination ability of the subject, is easily influenced by factors such as attention, understanding ability, reaction time, psychological state and the like, and is not suitable for infants, the elderly, patients with consciousness disturbance and people who cannot be effectively coordinated with the test. In order to overcome the limitations of subjective hearing tests, objective hearing test techniques have been developed. The otoacoustic emission (Otoacoustic Emission, OAE) detection technology can objectively reflect the inner ear and outer hair cell function state by recording acoustic signals generated by the active activities of the outer hair cells of the cochlea in the external auditory canal, and is widely applied to neonatal hearing screening and partial clinical application. However, otoacoustic emission signals need to pass through the middle ear conduction system during propagation from the inner ear to the external auditory canal, with amplitude and signal-to-noise ratio susceptible to significant changes in middle ear conditions. For example, when there are conditions of reduced tympanic membrane compliance, limited ossicular chain activity, negative ear canal pressure, or middle ear effusion, even if the inner ear is functioning properly, significant attenuation or inability to elicit otoacoustic emission signals can result, producing false negative results. On the other hand, middle ear function detection technology, especially wideband middle ear impedance detection technology, (Wideband Tympanometry, WBT or Wideband Admittance/absorpance, WBA) can evaluate the transmission characteristics of middle ear to acoustic energy in a wider frequency range, reflect the energy absorption, reflection and resonance characteristics of middle ear system, and has important significance for identifying middle ear conduction abnormality. However, the broadband middle ear impedance detection mainly reflects the middle ear mechanical conduction state, and is difficult to directly describe the active nonlinear function of the outer hair cells of the inner ear, and a simple one-to-one correspondence between the detection result and the hearing threshold value does not exist. In the prior art, although a hearing evaluation method based on otoacoustic emission and a middle ear function detection method based on acoustic immittance or broadband middle ear impedance exist respectively, most technical schemes only use single detection information and cannot perform joint modeling on middle ear conduction characteristics and inner ear active functions in the same detection flow, so that contradiction between middle ear abnormal interference and hearing threshold prediction accuracy is difficult to solve simultaneously. Therefore, a hearing test system that can integrate multiple sources of information from the middle ear and the inner ear is needed to improve accuracy, robustness, and clinical applicability of objective hearing threshold prediction. Disclosure of Invention Aiming at the defects of the prior art, the invention provides a wideband middle ear impedance and otoacoustic emission combined hearing threshold prediction method, which overcomes the defects of the prior art, and the problems of insufficient accuracy and insensitivity to middle ear abnormality caused by only relying on a single detection means in the prior art are overcome by synchronously acquiring middle ear conduction characteristics and inner ear outer hair cell function information in the same detection flow and constructing a combined characteristic input model for analysis. In order to achieve the above purpose, the invention is realized by the following technical scheme: the invention provides a method for predicting the hearing threshold of a combination of broadband middle ear impedance and otoacoustic emission, which comprises the following steps: determining a detection mode according to user selection or system preset rules, and initializing detection parameters corresponding to the detection mode; after finishing the steps of selecting the detection mode and initializing parameters, performing auditory canal sea