Adaptive Audiometry Sequencing for Faster Threshold Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audiometry methods struggle to efficiently and accurately determine hearing thresholds across a wide audience and in cases of severe hearing loss, particularly due to over-stimulation and cochlear dead zones.
Innovation Solution
A method utilizing a Gaussian process to determine sound sequences, adjusting intensity and frequency based on patient responses, incorporating an exponential and linear kernel to optimize threshold estimation, and employing an electronic device for audiometry testing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional audiometry methods are used to determine hearing thresholds, then the measurement can be obtained, but the process is time-consuming and less accurate for severe hearing loss
Solution Approach 1:
The system uses the patient's responses to sound stimuli to dynamically adjust the testing sequence. The Gaussian process model continuously updates the audiogram estimation based on observed responses, and this feedback drives the selection of subsequent sounds to present, optimizing both speed and accuracy of threshold determination
Solution Approach 2:
The system changes the parameters of sound stimuli (frequency and intensity) dynamically based on the Gaussian process model's uncertainty estimation. Sounds are selected to maximize information gain by targeting frequencies and intensities where the model is most uncertain, thereby efficiently converging on the true hearing threshold
2Productivity
If sound intensity is rapidly increased to determine hearing thresholds, then the testing speed improves, but over-stimulation occurs causing inaccurate results
Solution Approach 1:
The sound intensity progression is dynamic rather than fixed. The system adjusts the rate of intensity increase based on patient responses and the Gaussian process model's uncertainty, slowing down when over-stimulation risk is detected and accelerating when confidence is high, thereby maintaining both speed and accuracy
Solution Approach 2:
Patient responses provide continuous feedback that regulates sound intensity progression. When the model detects patterns suggesting over-stimulation (e.g., unexpected hearing at high intensities), it adjusts subsequent sound selection to avoid further rapid intensity increases, ensuring reliable threshold determination
3Adaptability or versatility
If a comprehensive sound range is tested to cover all frequencies, then the audiogram completeness improves, but the testing complexity and time increase
Solution Approach 1:
The system dynamically changes which frequency and intensity parameters are tested based on the Gaussian process model's uncertainty distribution. Rather than uniformly testing all frequencies, it adaptively selects parameters that will most reduce overall uncertainty in the audiogram estimation
Solution Approach 2:
The comprehensive frequency range is effectively segmented into regions of high and low uncertainty. The system focuses testing efforts on frequency regions where the model is uncertain, while relying on model interpolation for regions where data already exists, thereby reducing overall testing complexity while maintaining comprehensive coverage
Data Source
AI summary
A method for audiometry testing of an ear comprising the repetition of the following steps:Determination (S30) of a second sound with a second frequency and a second intensity from a Gaussian process,Determination (S40), if present, of a third sound already applied to the ear having a third intensity, lower than the second intensity and higher than a second threshold and having a third frequency, the difference between the third frequency and the second frequency being lower than a frequency threshold, the third intensity being maximum among all the sounds whose frequency has a difference with the second frequency lower than the frequency threshold,Application (S50) to the ear:A sound with the second frequency and the third intensity increased by a predefined increment, if the second intensity is greater than the second threshold, and if a third sound exists,Otherwise, second sound.