Detection of cognitive impairment using speech feature distribution
A method, computer system, and a computer program product for speech feature distribution is provided. The present invention may include receiving two or more speech samples from a user. The present invention may include dividing the two or more speech samples into a a plurality of pause segments. The present invention may include determining a pause duration for each of the plurality of pause segments. The present invention may include determining a distribution of pause durations. The present invention may include determining a distance between the distribution of pause durations.
1. A method for speech feature distribution, the method comprising:
receiving two or more speech samples from a user, wherein a first speech sample corresponds to a first task, wherein a second speech sample corresponds to a second task, and wherein the first task and the second task are each one of at least two or more tasks selected by a medical professional within a speech user interface, wherein each of the at least two or more tasks are retrieved from a knowledge corpus and are of different cognitive difficulty levels;
dividing at least the first speech sample and the second speech sample into a plurality of pause segments using a Voice Activity Detection (VAD) feature of one or more speech-to-text engines, wherein the plurality of pause segments are below a sound intensity level threshold set or adjusted by the medical professional in the speech user interface;
determining a first pause duration distribution from the plurality of pause segments of the first speech sample;
determining a second pause duration distribution from the plurality of pause segments of the second speech sample;
determining a distance between the first pause duration distribution and the second pause duration distribution; and
comparing the distance with a corresponding distance between pause duration distributions from a second user for a same at least two or more tasks, wherein the second user has a cognitive impairment.
2. The method of claim 1 , further comprising:
determining a probability in which the user is suffering from the cognitive impairment of the second user based on the comparing of the distance between each of the distributions of the pause durations of the two or more speech samples received for the user and the second user, wherein the distributions of the pause durations for the second user are retrieved from the knowledge corpus.
3. The method of claim 2 , further comprising:
displaying one or more recommendations to the medical professional within the speech user interface based on the probability in which the user is suffering from the cognitive impairment, wherein the one or more recommendations are based on an efficacy of treatments for the second user stored in the knowledge corpus.
4. The method of claim 1 , wherein selecting the first task and the second task from the knowledge corpus further comprises:
retrieving at least two lists of similar individuals who have performed a same task, based on data input received from the medical professional with respect to the user, the data input being entered by the user in the speech user interface;
displaying the at least two lists of similar individuals to the medical professional in the speech user interface;
displaying the first task and the second task to the medical professional in the speech user interface based on the similar individuals selected by the medical professional from the at least two lists; and
adjusting the first task and the second task based on a modification of the medical professional.
5. The method of claim 1 , further comprising:
determining a probability in which the user is suffering from at least one cognitive impairment of a plurality of cognitive impairments using one or more machine learning models, wherein the one or more machine learning models are trained according to a holdout or cross validation method and are based on a plurality of speech features and plurality of data stored in the knowledge corpus, wherein an input received by the one or more machine learning models is selected from the group consisting of the distance between the distribution of the pause durations for the user, a plurality of additional speech features determined from the two or more speech samples of the user, and a plurality of user data inputted by the medical professional.
6. The method of claim 5 , further comprising:
determining, using the one or more machine learning models, which of the plurality of cognitive impairments the user is suffering from;
determining, using the one or more machine learning models, a stage of the at least one cognitive impairment; and
providing, in the speech user interface, one or more recommendations to the user for the at least one cognitive impairment, wherein the one or more recommendations are determined using the one or more machine learning models utilizing as input at least, one or more of, which of the plurality of cognitive impairments the user is suffering from, the stage of the at least one cognitive impairment, and data stored in the knowledge corpus.
7. The method of claim 1 , wherein the distance between each of the distributions for each of the pause durations is determined utilizing one or more distribution techniques, wherein the one or more distribution techniques are selected based on the tasks selected by the medical professional, a highest area under a curve (AUC), and a lowest p value.
8. A computer system for speech feature distribution, comprising:
one or more processors, one or more computer-readable memories, one or more computer-readable tangible storage medium, and program instructions stored on at least one of the one or more tangible storage medium for execution by at least one of the one or more processors via at least one of the one or more memories, wherein the computer system is capable of performing a method comprising:
receiving two or more speech samples from a user, wherein a first speech sample corresponds to a first task, wherein a second speech sample corresponds to a second task, and wherein the first task and the second task are each one of at least two or more tasks selected by a medical professional within a speech user interface, wherein each of the at least two or more tasks are retrieved from a knowledge corpus and are of different cognitive difficulty levels;
dividing at least the first speech sample and the second speech sample into a plurality of pause segments using a Voice Activity Detection (VAD) feature of one or more speech-to-text engines, wherein the plurality of pause segments are below a sound intensity level threshold set or adjusted by the medical professional in the speech user interface;
determining a first pause duration distribution from the plurality of pause segments of the first speech sample;
determining a second pause duration distribution from the plurality of pause segments of the second speech sample;
determining a distance between the first pause duration distribution and the second pause duration distribution; and
comparing the distance with a corresponding distance between pause duration distributions from a second user for a same at least two or more tasks, wherein the second user has a cognitive impairment.
9. The computer system of claim 8 , further comprising:
determining a probability in which the user is suffering from the cognitive impairment of the second user based on the comparing of the distance between each of the distributions of the pause durations of the two or more speech samples received for the user and the second user, wherein the distributions of the pause durations for the second user are retrieved from the knowledge corpus.
10. The computer system of claim 9 , further comprising:
displaying one or more recommendations to the medical professional within the speech user interface based on the probability in which the user is suffering from the cognitive impairment, wherein the one or more recommendations are based on an efficacy of treatments for the second user stored in the knowledge corpus.
11. The computer system of claim 8 , wherein selecting the first task and the second task from the knowledge corpus further comprises:
retrieving at least two lists of similar individuals who have performed a same task, based on data input received from the medical professional with respect to the user, the data input being entered by the user in the speech user interface;
displaying the at least two lists of similar individuals to the medical professional in the speech user interface;
displaying the first task and the second task to the medical professional in the speech user interface based on the similar individuals selected by the medical professional from the at least two lists; and
adjusting the first task and the second task based on a modification of the medical professional.
12. The computer system of claim 8 , further comprising:
determining a probability in which the user is suffering from at least one of a plurality of cognitive impairments using one or more machine learning models, wherein the one or more machine learning models are trained according to a holdout or cross validation method and are based on a plurality of speech features and plurality of data stored in the knowledge corpus, wherein an input received by the one or more machine learning models is selected from the group consisting of the distance between the distribution of the pause durations for the user, a plurality of additional speech features determined from the two or more speech samples of the user, and a plurality of user data inputted by the medical professional.
13. The computer system of claim 12 , further comprising:
determining, using the one or more machine learning models, which of the plurality of cognitive impairments the user is suffering from;
determining, using the one or more machine learning models, a stage of the at least one cognitive impairment; and
providing, in the speech user interface, one or more recommendations to the user for the at least one cognitive impairment, wherein the one or more recommendations are determined using the one or more machine learning models utilizing as input at least, one or more of, which of the plurality of cognitive impairments the user is suffering from, the stage of the at least one cognitive impairment, and data stored in the knowledge corpus.
14. The computer system of claim 8 , wherein the distance between each of the distributions for each of the pause durations is determined utilizing one or more distribution techniques, wherein the one or more distribution techniques are selected based on the tasks selected by the medical professional, a highest area under a curve (AUC), and a lowest p value.
15. A computer program product for speech feature distribution, comprising:
one or more non-transitory computer-readable storage media and program instructions stored on at least one of the one or more tangible storage media, the program instructions executable by a processor to cause the processor to perform a method comprising:
receiving two or more speech samples from a user, wherein a first speech sample corresponds to a first task, wherein a second speech sample corresponds to a second task, and wherein the first task and the second task are each one of at least two or more tasks selected by a medical professional within a speech user interface, wherein each of the at least two or more tasks are retrieved from a knowledge corpus and are of different cognitive difficulty levels;
dividing at least the first speech sample and the second speech sample into a plurality of pause segments using a Voice Activity Detection (VAD) feature of one or more speech-to-text engines, wherein the plurality of pause segments are below a sound intensity level threshold set or adjusted by the medical professional in the speech user interface;
determining a first pause duration distribution from the plurality of pause segments of the first speech sample;
determining a second pause duration distribution from the plurality of pause segments of the second speech sample;
determining a distance between the first pause duration distribution and the second pause duration distribution; and
comparing the distance with a corresponding distance between pause duration distributions from a second user for a same at least two or more tasks, wherein the second user has a cognitive impairment.
16. The computer program product of claim 15 , further comprising:
determining a probability in which the user is suffering from the cognitive impairment of the second user based on the comparing of the distance between each of the distributions of the pause durations of the two or more speech samples received for the user and the second user, wherein the distributions of the pause durations for the second user are retrieved from the knowledge corpus.
17. The computer program product of claim 16 , further comprising:
displaying one or more recommendations to the medical professional within the speech user interface based on the probability in which the user is suffering from the cognitive impairment, wherein the one or more recommendations are based on an efficacy of treatments for the second user stored in the knowledge corpus.
18. The computer program product of claim 15 , wherein selecting the first task and the second task from the knowledge corpus further comprises:
retrieving at least two lists of similar individuals who have performed a same task, based on data input received from the medical professional with respect to the user, the data input being entered by the user in the speech user interface;
displaying the at least two lists of similar individuals to the medical professional in the speech user interface;
displaying the first task and the second task to the medical professional in the speech user interface based on the similar individuals selected by the medical professional from the at least two lists; and
adjusting the first task and the second task based on a modification of the medical professional.
19. The computer program product of claim 15 , further comprising:
determining a probability in which the user is suffering from at least one of a plurality of cognitive impairments using one or more machine learning models, wherein the one or more machine learning models are trained according to a holdout or cross validation method and are based on a plurality of speech features and plurality of data stored in the knowledge corpus, wherein an input received by the one or more machine learning models is selected from the group consisting of the distance between the distribution of the pause durations for the user, a plurality of additional speech features determined from the two or more speech samples of the user, and a plurality of user data inputted by the medical professional.
20. The computer program product of claim 15 , wherein the distance between each of the distributions for each of the pause durations is determined utilizing one or more distribution techniques, wherein the one or more distribution techniques are selected based on the tasks selected by the medical professional, a highest area under a curve (AUC), and a lowest p value.