Methods for remote visual identification of congestive heart failures

1. A method for remote diagnosis of a congestive heart failure in humans, comprising:

receiving a facial image of a patient, wherein the facial image is retrieved from a data store and captured from the patient;

extracting, from the facial image, at least one facial feature indicative of a heart condition, wherein extracting further comprises generating numerical descriptors of facial attributes for the at least one facial feature, wherein the numerical descriptors provide a facial geometry of the at least one facial feature;

classifying the extracted at least one facial feature using a classifier, wherein the classifier maps a plurality of candidate facial features to a plurality of scores indicating a stage of the heart condition; and

determining, using the classifier, a positive diagnosis and the stage of the heart condition of the patient based on the plurality of scores of the plurality of candidate facial features.

2. The method of claim 1 , further comprising:

pre-processing of the facial image to standardize the facial image and detect a face within the facial image.

3. The method of claim 1 , further comprising:

comparing the at least one facial feature to sample facial features, wherein the sample facial features include data labels associated with respective sample facial features; and

selecting at least one matching sample facial feature that compares substantially closely to the at least one facial feature.

4. The method of claim 1 , wherein the classifier is trained to provide a confidence rating for each of the plurality of scores.

5. The method of claim 1 , wherein the at least one facial feature indicative of the heart condition is any one of: eyes, lips, skin, ear, forehead, facial muscle tone, size, shape, length, facial color, and cheeks.

6. The method of claim 1 , further comprising:

generating the classifier, wherein the classifier is generated for a unique type of heart disease.

7. The method of claim 6 , further comprising:

receiving a labeled dataset including a set of facial images, wherein each of the facial images is labeled depending on whether it represents a heart disease;

extracting, from each facial image in the set of facial images, at least one learning facial feature indicative of the heart condition; and

feeding the extracted facial features into a deep neural network to produce a trained model, wherein the classifier is generated and ready when the trained model includes enough facial features processed by the deep neural network.

8. The method of claim 7 , wherein the deep neural network implements machine learning techniques including at least: a semi-supervised machine learning algorithm.

9. The method of claim 7 , further comprising:

feeding back a score of the determined positive diagnosis to the trained model.

10. The method of claim 1 , wherein the facial geometry of the at least one facial feature includes at least one of: size, shape, position, length, color, and relationship between the at least one facial feature.

11. A non-transitory computer readable medium having stored thereon instructions for a processing circuitry to execute the process, the process comprising:

receiving a facial image of a patient, wherein the facial image is retrieved from a data store and captured from the patient;

extracting, from the facial image, at least one facial feature indicative of a heart condition, wherein extracting further comprises generating numerical descriptors of facial attributes for the at least one facial feature, wherein the numerical descriptors provide a facial geometry of the at least one facial feature;

classifying the extracted at least one facial feature using a classifier, wherein the classifier maps a plurality of candidate facial features to a plurality of scores indicating a stage of the heart condition; and

determining, using the classifier, a positive diagnosis and the stage of the heart condition of the patient based on the plurality of scores of the plurality of candidate facial features.

12. A system for remote diagnosis of a congestive heart failure in humans, comprising:

a processing circuitry; and

a memory, the memory containing instructions that, when executed by the processing circuitry, configure the system to:

receive a facial image of a patient, wherein the facial image is retrieved from a data store and captured from the patient;

extract, from the facial image, at least one facial feature indicative of a heart condition, wherein extracting further comprises generating numerical descriptors of facial attributes for the at least one facial feature, wherein the numerical descriptors provide a facial geometry of the at least one facial feature;

classify the extracted at least one facial feature using a classifier, wherein the classifier maps a plurality of candidate facial features to a plurality of scores indicating a stage of the heart condition; and

determine, using the classifier, a positive diagnosis and the stage of the heart condition of the patient based on the plurality of scores of the plurality of candidate facial features.

13. The system of claim 12 , wherein the system is further configured to:

pre-process the facial image to standardize the facial image and detect a face within the facial image.

14. The system of claim 12 , wherein the system is further configured to:

compare the at least one facial feature to sample facial features, wherein the sample facial features include data labels associated with respective sample facial features; and

select at least one matching sample facial feature that compares substantially closely to the at least one facial feature.

15. The system of claim 12 , wherein the classifier is trained to provide a confidence rating for each of the plurality of scores.

16. The system of claim 12 , wherein the at least one facial feature indicative of the heart condition is any one of: eyes, lips, skin, ear, forehead, facial muscle tone, size, shape, length, facial color, and cheeks.

17. The system of claim 12 , wherein the system is further configured to:

generate the classifier, wherein the classifier is generated for a unique type of heart disease.

18. The system of claim 17 , wherein the system is further configured to:

receive a labeled dataset including a set of facial images, wherein each of the facial images is labeled depending on whether it represents a heart disease;

extract, from each facial image in the set of facial images, at least one learning facial feature indicative of the heart condition; and

feed the extracted facial features into a deep neural network to produce a trained model, wherein the classifier is generated and ready when the trained model includes enough facial features processed by the deep neural network.

19. The system of claim 18 , wherein the deep neural network implements machine learning techniques including at least: a semi-supervised machine learning algorithm.

20. The system of claim 18 , wherein the system is further configured to:

feedback a score of the determined positive diagnosis to the trained model.