Diagnosis and treatment of speech and language pathologies by speech to text and natural language processing

1. A computerized method for assessing a speech and/or lingual quality of a subject, and treating and/or diagnosing a speech and/or lingual related pathology, the method comprising:

providing a content-containing stimulus to a user;

recording the user's vocal response to the stimulus and/or to instructions related thereto;

extracting and/or measuring and/or computing at least one prosodic parameter and at least one acoustic parameter from the user's recorded vocal response by a processing unit, wherein the at least one acoustic parameter comprises a temporal parameter which comprises the user's response time;

identifying the words in the user's vocal response by the processing unit and transforming the user's vocal response into a transformed text section, based on the identified words in the user's vocal response;

comparing the transformed text section to a predetermined text section via the processing unit, which represents the user's expected response;

computing by the processing unit, an output signal indicative of at least one speech and/or lingual quality of the user, which comprises computing a multi parametric vector, based at least on data resulted from the texts comparison, the at least one extracted and/or measured and/or computed prosodic parameter and acoustic parameter; and

detecting a speech and/or lingual related pathology of the subject suffering from said pathology based upon the output signal, wherein the speech and/or lingual related pathology is selected from a group consisting of aphasia, ALS, Parkinson, Alzheimer's and stuttering;

wherein computing an output signal indicative of the user's speech quality, comprises applying natural language processing (NLP) algorithms,

wherein the natural language processing (NLP) algorithms comprise machine learning algorithms,

wherein the machine learning algorithms comprise decision trees, neural networks, statistical models, or a combination thereof, and

wherein the statistical models comprise assigning a weight value to each of the parameters or sets of parameters and wherein the weight value is displayed on a display.

2. The method of claim 1 , wherein computing the output signal comprises assigning a speech quality score representing deviation from a norm.

3. The method of claim 1 , wherein computing the output signal comprises assigning a lingual quality score representing deviation from a norm.

4. The method of claim 1 , wherein the at least one speech quality comprises a sub-set of speech qualities.

5. The method of claim 1 , wherein the stimulus provided to the user comprises a text, image, video, voice stimulus, or any combination thereof.

6. The method of claim 1 , wherein computing the level of similarity between the transformed text section and the predetermined text section is determined based on text content parameters.

7. The method of claim 1 , wherein the at least one temporal parameter further comprises the user's speech rate, length of prolongation, length of blocking, lengthy of pauses between words, or any combination thereof.

8. The method of claim 1 , wherein the at least one acoustic parameter relates to redundancies, interjections, speech intonation, tone, stress, rhythm, intensity, pitch, loudness, mental state or any combination thereof.

9. The method of claim 1 , wherein the at least one linguistic parameter relates to vocabulary, phonology, morphology, syntax, semantics, lexical status, pragmatics or any combination thereof.

10. The method of claim 1 , wherein parameters related to the multi parametric vector comprise phonetic transcription, part of speech, semantic class, intensity, pitch, loudness, intonation, tone, stress, rhythm number of correct words, percentage of time associated with correct words, degree of acoustic similarity, degree of semantic similarity, or any combination thereof.

11. The method of claim 1 , wherein parameters related to the multi parametric vector comprise lingual and acoustic parameters.

12. The method of claim 11 , wherein the lingual parameters relate to comprehension, pronunciation, planning and/or organization of correct grammar, pragmatic skills of communication, or any combination thereof.

13. A processing unit for assessing a speech and/or lingual quality of a subject, and treating and/or diagnosing a speech and/or lingual related pathology, the processing unit comprising:

a vocal input sub-unit executing a code configured to receive a recorded vocal sample created by a user's response to content-containing stimulus and/or to instructions related thereto;

a speech processing sub-unit executing a code configured to extract and/or measure and/or compute at least one prosodic parameter and at least one acoustic parameter relating to the user's vocal response, wherein the at least one acoustic parameter comprises a temporal parameter which comprises the user's response time;

a vocal-to-text sub-unit executing a code configured to transform the user's verbal response into a transformed text section;

a comparison sub-unit executing a code configured to compare between the transformed text section and a predetermined text section, which represents the user's expected response; and

an output unit executing a code configured to:

compute an output signal indicative of at least one speech and/or lingual quality of the user, which comprises computing a multi parametric vector, based at least on data resulted from the texts comparison, the at least one extracted and/or measured and/or computed prosodic parameter and the at least one acoustic parameter; and

detect a speech and/or lingual related pathology of the subject suffering from said pathology based upon the output signal wherein the speech and/or lingual related pathology is selected from a group consisting of aphasia, ALS, Parkinson, Alzheimer's and stuttering;

wherein computing an output signal indicative of the user's speech quality, comprises applying natural language processing (NLP) algorithms,

wherein the natural language processing (NLP) algorithms comprise machine learning algorithms,

wherein the machine learning algorithms comprise decision trees, neural networks, statistical models, or a combination thereof, and

wherein the statistical models comprise assigning a weight value to each of the parameters or sets of parameters and wherein the weight value is displayed on a display.

14. A system for assessing a speech and/or lingual quality of a subject, and treating and/or diagnosing a speech and/or lingual related pathology, the system comprising:

a prompter configured to provide a content-containing stimulus to a user;

a recorder configured to record a vocal sample created by the user's response to content-containing stimulus and/or to instructions related thereto; and

a processing unit comprising:

a vocal input sub-unit executing a code configured to receive a recorded vocal sample created by a user's response to content-containing stimulus and/or to instructions related thereto;

a speech processing sub-unit executing a code configured to extract and/or measure and/or compute at least one prosodic parameter and at least one acoustic parameter relating to the user's vocal response, wherein the at least one acoustic parameter comprises a temporal parameter which comprises the user's response time;

a vocal-to-text sub-unit executing a code configured to transform the user's verbal response into a transformed text section;

a comparison sub-unit executing a code configured to compare between the transformed text section and a predetermined text section, which represents the user's expected response; and

an output sub-unit executing a code configured to:

compute an output signal indicative of at least one speech and/or lingual quality of the user, which comprises computing a multi parametric vector, based at least on data resulted from the texts comparison, the at least one extracted and/or measured and/or computed prosodic parameter and the at least one acoustic parameter; and

detect a speech and/or lingual related pathology of the subject suffering from said pathology based upon the output signal wherein the speech and/or lingual related pathology is selected from a group consisting of aphasia, ALS, Parkinson, Alzheimer's and stuttering;

wherein computing an output signal indicative of the user's speech quality, comprises applying natural language processing (NLP) algorithms,

wherein the natural language processing (NLP) algorithms comprise machine learning algorithms,

wherein the machine learning algorithms comprise decision trees, neural networks, statistical models, or a combination thereof, and

wherein the statistical models comprise assigning a weight value to each of the parameters or sets of parameters and wherein the weight value is displayed on a display.