Digital image screening and/or diagnosis using artificial intelligence

1. A device configured to capture an image of a patient, the device comprising:

a camera configured to capture the image of the patient;

at least one processor; and

at least one system memory encoding instructions which, when executed by the at least one processor, cause the device to:

capture contextual data associated with the image, wherein the contextual data includes a process by which a caregiver has selected images for storage and discard, and wherein the contextual data further includes manipulation data relating to manipulations of the image performed by the caregiver;

provide feedback to the caregiver relating to a quality of the image as the image is captured, wherein the feedback includes at least one recommendation to improve the quality of the image;

allow the caregiver to select the image to be stored;

forward the image and the contextual data to a remote server for processing using artificial intelligence;

receive a proposed diagnosis and an accuracy indicator indicating an accuracy of the proposed diagnosis determined by the artificial intelligence on the remote server, the proposed diagnosis being based upon the image and the contextual data, and the accuracy indicator being a score within a range having a minimum score and a maximum score; and

receive a recommendation determined by the artificial intelligence on the remote server, the recommendation including a suggestion to perform one or more additional tests to be done by the device to increase the accuracy indicator of the proposed diagnosis, the one or more additional tests including taking additional images of the patient using the camera of the device and providing additional contextual data to increase the accuracy indicator of the proposed diagnosis.

2. The device of claim 1 , wherein the minimum score indicates less certainty and the maximum score indicates more certainty regarding the potential accuracy of the proposed diagnosis.

3. The device of claim 1 , wherein the proposed diagnosis is a multi-class classification.

4. The device of claim 1 , wherein the at least one system memory encodes further instructions which, when executed by the at least one processor, cause the device to receive a second diagnosis from a referral service.

5. A device configured to process a fundus image of a patient, the device comprising:

at least one processor; and

at least one system memory encoding instructions which, when executed by the at least one processor, cause the device to:

receive an image of a fundus of the patient, the image of the fundus captured by a second device;

receive contextual data associated with the image, wherein the contextual data includes a process by which a caregiver has selected images for storage and discard, and wherein the contextual data further includes manipulation data relating to manipulations of the image performed by the caregiver;

receive feedback relating to a quality of the image as the image is captured, wherein the feedback includes at least one recommendation to improve the quality of the image;

use artificial intelligence to develop a proposed diagnosis based upon the image and the contextual data;

calculate an accuracy indicator indicating an accuracy of the proposed diagnosis, the accuracy indicator being within a range having a minimum score and a maximum score; and

suggest one or more additional tests using the artificial intelligence, the one or more additional tests to be done by the second device to increase the accuracy indicator of the proposed diagnosis including taking additional images by the second device and providing additional contextual data to increase the accuracy indicator of the proposed diagnosis.

6. The device of claim 5 , wherein the minimum score indicates less certainty and the maximum score indicates more certainty regarding the potential accuracy of the proposed diagnosis.

7. The device of claim 5 , wherein the proposed diagnosis is a multi-class classification.

8. The device of claim 5 , wherein the at least one system memory encodes further instructions which, when executed by the at least one processor, cause the device to forward the image to a referral service for a second diagnosis.

9. A method for diagnosing a disease associated with a fundus of a patient, the method comprising:

receiving an image of the fundus of the patient, the image of the fundus captured by a device;

receiving contextual data associated with the image, wherein the contextual data includes a process by which a caregiver has selected images for storage and discard, and wherein the contextual data further includes manipulation data relating to manipulations of the image performed by the caregiver;

receiving feedback relating to a quality of the image as the image is captured, wherein the feedback includes at least one recommendation to improve the quality of the image;

using artificial intelligence to develop a proposed diagnosis based upon the image and the contextual data;

calculating an accuracy indicator indicating an accuracy of the proposed diagnosis, the accuracy indicator being within a range having a minimum score and a maximum score; and

suggesting one or more additional tests using the artificial intelligence, the one or more additional tests to be done by the device to increase the accuracy indicator of the proposed diagnosis including taking additional images by the device and providing additional contextual data to increase the accuracy indicator of the proposed diagnosis.

10. The method of claim 9 , wherein the minimum score indicates less certainty and the maximum score indicates more certainty regarding the potential accuracy of the proposed diagnosis.

11. The method of claim 9 , forwarding the image to a referral service for a second diagnosis.