Cough detection system

Patent metadata

Field	Value
Publication number	US-12159646-B2
Application number	US-202117188462-A
Country	US
Kind code	B2
Filing date	Mar 1, 2021
Priority date	Mar 1, 2021
Publication date	Dec 3, 2024
Grant date	Dec 3, 2024

Abstract

Methods and systems for a cough detection system are provided. The methods and systems include operations comprising: receiving an utterance comprising a cough from a user; determining one or more environmental factors, one or more health related factors, or both associated with the user; processing the utterance with a machine learning technique to generate an illness prediction, the machine learning technique being trained to establish a relationship between a plurality of training cough utterances and a plurality of illnesses corresponding to the plurality of training cough utterances; applying a weight to the illness prediction based on the one or more environmental factors, health related factors, or both associated with the user; and triggering an alert representing the illness prediction based on the weighted illness prediction.

First claim

What is claimed is:

1. A method comprising:
receiving, by one or more processors, an utterance including a cough from a user;
in response to receiving the utterance, automatically determining, by the one or more processors, a set of environmental factors and a set of health related factors associated with the user, wherein the set of environmental factors includes at least one of: a spectral density associated with the utterance, an intensity of the utterance, an amount of ambient noise present in the utterance, a measure of humidity in an area in which the utterance was captured, an indication of whether a mask was worn by the user when the utterance was captured, or an indication of whether the utterance was captured indoors or outdoors;
in response to determining the set of environmental factors and the set of health related factors, automatically generating, by the one or more processors, a sound clip of the utterance including only the cough;
inputting the sound clip into a machine learning model to generate an illness prediction, wherein the machine learning model is trained to establish a relationship between a plurality of training cough utterances and a plurality of illnesses corresponding to the plurality of training cough utterances;
applying a weight to the illness prediction based on at least one environmental factor of the set of environmental factors or at least one health related factor of the set of health related factors;
triggering an alert representing the illness prediction based on the weighted illness prediction, including:
adding the alert to a patient record associated with the user,
storing the patient record in a database, and
transmitting the alert to a client device associated with the user;

generating a three-dimensional image including a location of the utterance and a set of individuals located proximate the location of the utterance; and
displaying the three-dimensional image on the client device.

2. The method of claim 1 , wherein:
the client device includes a microphone,
the utterance is captured by the microphone, and
the client device transmits the utterance to a server to generate the illness prediction.

3. The method of claim 1 , further comprising:
pre-processing the utterance; and
detecting a presence of the cough in the utterance.

4. The method of claim 1 , wherein the set of health related factors associated with the user includes a pre-existing health condition of the user, a medical history of the user, an indication of whether the user has a fever, or an indication of whether the user experiences fatigue.

5. The method of claim 1 , wherein the machine learning model includes a neural network.

6. The method of claim 1 , further comprising:
causing display, on the client device associated with the user, a survey including one or more health related questions;
receiving input from the user including responses to the survey, wherein the responses include the set of health related factors; and
causing the client device to capture, during completion of the survey, the sound clip.

7. The method of claim 6 , further comprising:
determining that the weighted illness prediction exceeds a specified threshold; and
in response to determining that the weighted illness prediction exceeds the specified threshold, causing the client device to notify the user that the responses and the sound clip are indicative of an illness.

8. The method of claim 1 , further comprising:
receiving the utterance from a microphone in a public space; and
applying a plurality of additional weights to the illness prediction, wherein:
each weight of the plurality of additional weights is associated with a different distance from the location of the utterance,
a first weight of the plurality of additional weights is applied to the illness prediction in association with a first region that is within a first distance to the location of the utterance to generate a second weighted illness prediction,
a second weight of the plurality of additional weights is applied to the illness prediction in association with a second region that is within a second distance to the location of the utterance to generate a third weighted illness prediction, and
the second distance is greater than the first distance.

9. The method of claim 8 , further comprising:
identifying a first client device that was within the first region when the utterance was captured;
determining that the second weighted illness prediction exceeds a specified threshold; and
in response to determining that the second weighted illness prediction exceeds the specified threshold, causing the first client device to notify a user of the first client device of a possible exposure to an illness associated with the illness prediction.

10. The method of claim 9 , further comprising:
identifying a second client device that was within the second region when the utterance was captured;
determining that the third weighted illness prediction associated with the second region fails to exceed the specified threshold; and
in response to determining that the third weighted illness prediction associated with the second region fails to exceed the specified threshold, preventing the second client device from notifying a user of the second client device of the possible exposure to the illness.

11. The method of claim 8 , further comprising:
identifying a first client device that entered the first region after a given period of time has elapsed from when the utterance was captured;
reducing the second weighted illness prediction associated with the first region based on the given period of time;
determining that the reduced second weighted illness prediction associated with the first region exceeds a specified threshold; and
in response to determining that the reduced second weighted illness prediction associated with the first region exceeds the specified threshold, causing the first client device to notify a user of the first client device of a possible exposure to an illness associated with the illness prediction.

12. The method of claim 8 , further comprising:
performing a first adjustment to the second and the third weighted illness predictions associated with the first and second regions based on whether the public space is indoors or outdoors; and
performing a second adjustment to the second and the third weighted illness predictions associated with the first and second regions based on whether a mask was worn by the user.

13. The method of claim 8 , further comprising receiving input from client devices within the public space opting into receiving possible exposure notifications of an illness associated with the illness prediction.

14. The method of claim 1 , further comprising:
receiving a request from the user to conduct a test to detect an infection; and
selecting a testing kit of a plurality of testing kits based on the weighted illness prediction, wherein:
each testing kit of the plurality of testing kits is associated with a different level of accuracy, and
triggering the alert includes transmitting a notification to the client device of the user that identifies the selected testing kit that the user is authorized to use to conduct the test.

15. The method of claim 1 , further comprising:
determining that the illness prediction corresponds to a specific type of illness associated with a follow up; and
communicating to a client device of a health care professional an identification of the user and the weighted illness prediction.

16. The method of claim 1 , further comprisin

Assignees

• Express Scripts Strategic Dev Inc

Inventors

• Ciliberti John
• Zheleznyak Anatoly

Classifications

• G06N3/0442

  characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU] · CPC title

• G06N3/094

  Adversarial learning · CPC title

• G06N3/0455

  Auto-encoder networks; Encoder-decoder networks · CPC title

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title

• G06N3/0475

  Generative networks · CPC title