Granted Patent B2

US 12,212,929 · App. 17/802,983 · Granted Jan 28, 2025

Inventors: Sarah Verhulst (Ghent, BE); Fotios Drakopoulos (Ghent, BE); Arthur Van Den Broucke (Knokke-Heist, BE); Sarineh Keshishzadeh (Ghent, BE)

Assignee: UNIVERSITEIT GENT

H04R25/507H04R25/606

Abstract

Claims (24)

1. An artificial neural network-based method for converting an auditory stimulus to a processed auditory output, the method comprising the steps of:

a. generating a neural network-based personalized auditory response model based at least on the integrity of auditory nerve fibers, auditory nerve synapses or a combination of auditory nerve fibers and auditory nerve synapses in a subject, said personalized auditory response model representing an expected auditory response of said subject with an auditory profile to the auditory stimulus;

b. comparing the output of the personalized auditory response model with the output of a neural network-based desired auditory response model to determine an auditory response difference; wherein said neural network-based models consist of non-linear operations that make the auditory response difference differentiable;

c. using the determined differentiable auditory response difference to develop a neural network-based individualized auditory signal processing model of the subject, wherein the individualized auditory signal processing model is configured to minimize the determined auditory response difference; and,

d. applying the individualized neural-network-based auditory signal processing model to the auditory stimulus to produce a processed auditory output that matches the desired auditory response, when given as an input to the personalized auditory response model or to the subject.

2. The method of claim 1 , wherein the personalized auditory response model of step a. is determined by deriving and including a subject specific auditory profile or wherein the subject specific auditory profile is a subject specific auditory damage profile.

3. The method of claim 1 , wherein the desired auditory response is the response from a normal-hearing subject or a response with enhanced features.

4. The method of claim 1 , wherein the desired auditory response model and the personalized auditory response model comprise models of different stages of the auditory periphery.

5. The method of claim 1 , wherein a reference neural network that describes a normal-hearing auditory periphery is used as the desired auditory response model; wherein a corresponding hearing-impaired neural network is used as the personalized auditory response model; and wherein the individualized auditory signal processing model is a signal processing neural network model trained to process the auditory input and compensate for the degraded output of the hearing-impaired model, when connected to the input of the hearing-impaired model or the subject.

6. The method of claim 1 , wherein a reference neural network that simulates augmented hearing perception and/or ability of a normal-hearing listener is used as the desired auditory response model; wherein a corresponding normal-hearing or hearing-impaired neural network is used as the personalized auditory response model; and wherein the individualized auditory signal processing model is a signal processing neural network model trained to process the auditory input and provide an augmented auditory response.

7. The method of claim 1 , wherein the individualized auditory signal processing model is trained to minimize a specific auditory response difference metric, such as the absolute or squared difference between the two auditory response models at several or all tonotopic frequencies.

8. The method of claim 1 , wherein the processed auditory output is selected from a modified auditory stimulus which is devised to compensate for a hearing-impairment or yields augmented hearing.

9. The method of claim 1 , wherein the processed auditory output is selected from a modified auditory response corresponding to a specific processing stage along the auditory pathway, which can for example be used to stimulate auditory prostheses such as cochlear implants or deep brain implants.

10. The method of claim 1 , wherein the difference of auditory-nerve outputs of a normal-hearing and a hearing-impaired periphery is minimized; or wherein the difference between simulated auditory brainstem and/or cortical responses, expressed in the time or frequency domain, is minimized.

11. The method of claim 1 , wherein a task-optimized speech ‘back-end’ which simulates the performance of listeners in different tasks is connected to the outputs of the auditory response models, also referred to as ‘front-ends’; and wherein the outputs of the back-end are used to determine and minimize the auditory response difference.

12. The method of claim 1 , for configuring an auditory device, wherein the auditory device is a cochlear implant or a wearable hearing aid.

13. The method of claim 1 , wherein the auditory stimulus is processed to an auditory output by a cochlear implant or by a wearable hearing aid.

14. An auditory device, comprising:

an input device configured to pick up an input sound wave from the environment and convert the input sound wave to an auditory stimulus;

a processing unit configured for performing the method of claim 1 to produce a processed auditory output; and,

an output device configured to produce the processed auditory output from the processing unit.

15. The method of claim 1 , wherein the generating of the neural network-based personalized auditory response model is further based on the integrity of inner hair cell damage, outer hair cell damage, or a combination of inner hair cell damage and outer hair cell damage in the subject.

16. The device of claim 14 , wherein the auditory device is a cochlear implant or a wearable hearing aid.

17. A data storage device comprising a computer program product for implementing, when executed on a processor, the method of claim 1 .

Assignments (1)

Priority Claims (1)

Continuity (1)

References Cited (19)