Healthcare object recognition, systems and methods
Healthcare object (HCO) discriminator systems and methods are presented. Systems can obtain a digital representation of a scene via a sensor interface. An HCO discriminator platform analyzes the digital representation to discriminate objects within the scene as being associated with a type of HCO or as being unrelated to a type of HCO. Once the HCO recognition platform determines that a type of HCO is relevant, it instantiates an actual HCO. The HCO can be routed to one or more destinations based on routing rules generated from a template or based on the manner in which the objects in the scene were discriminated.
1. A method for routing healthcare objects, the method comprising:
receiving, by at least one processor, a digital representation of a scene from at least one sensor;
deriving a set of discriminator characteristics via execution of one or more discriminator algorithms on the digital representation according to data modalities within the digital representation, the one or more discriminator algorithms stored in memory, the set of discriminator characteristics comprising a vector of discriminator characteristics, the vector of discriminator characteristics comprising multiple vector members, and each vector member corresponding to a different modality of data;
discriminating objects in the scene by recognizing an object as a healthcare object type based on the set of discriminator characteristics;
generating a notification from the recognized object according to the healthcare object type;
routing the notification to a destination device according to routing rules associated with the notification; and
rendering a user interface on a display of the destination device in response to receiving the notification.
2. The method of claim 1 , wherein the notification includes information obtained from the digital representation of the scene.
3. The method of claim 1 , further comprising determining physician information associated with the recognized object.
4. The method of claim 3 , wherein the destination device is a physician destination device associated with the determined physician information.
5. The method of claim 3 , wherein the physician information is determined according to at least one of a recognized face of a person in the digital representation of the scene or a recognized license plate number in the digital representation of the scene.
6. The method of claim 1 , wherein the notification is a first healthcare object type, the method further including:
generating a second health care object from a second recognized object according to a healthcare object type of the second recognized object.
7. The method of claim 6 , further comprising routing the second health care object to a second destination device according to routing rules associated with the second health care object.
8. The method of claim 1 , wherein the user interface is rendered on at least one of a cell phone or a browser device.
9. The method of claim 1 , further comprising requesting additional medically-relevant information via the user interface.
10. The method of claim 1 , further comprising instantiating a healthcare object, wherein instantiating the healthcare object includes provisioning the healthcare object with destination attributes.
11. The method of claim 10 , wherein the destination attributes include at least one of the routing rules or a destination address.
12. The method of claim 11 , wherein the destination address comprises a document object identifier.
13. The method of claim 1 , wherein the modality of data comprises one of image features, audio features, or OCRed text.
14. The method of claim 1 , wherein:
the set of discriminator characteristics comprises an N-tuple of characteristics; and
the N-tuple of characteristics comprises attribute names along with one or more corresponding values.
15. The method of claim 1 , wherein the one or more discriminator algorithms include at least one of SIFT, BRISK, GLOW, SURF, SLAM, vSLAM, a hidden markov model, optical character recognition, a support vector machine, or dynamic time warping.
16. The method of claim 1 , further comprising recognizing objects in the scene as non-healthcare objects.
17. The method of claim 1 , further comprising instantiating a healthcare object, wherein instantiating the healthcare object includes populating fields of a healthcare object template.
18. The method of claim 1 , further comprising instantiating a healthcare object, wherein instantiating the healthcare object includes causing a mechanical turk computer system to construct a mechanical turk task.
19. The method of claim 18 , further comprising populating fields of the healthcare object based on information resulting from the mechanical turk task.
20. The method of claim 1 , further comprising encrypting a healthcare object.