Maintaining a Discrete Data Representation that Corresponds to Information Contained in Free-Form Text

Abstract

A system includes a data record (such as an Electronic Medical Record (EMR)) and a user interface for modifying (e.g., storing data in) the data record. The data record includes both free-form text elements and discrete data elements. The user interface includes user interface elements for receiving free-form text data. In response to receiving free-form text data via the free-form text user interface elements, a suggested action is identified, such as a suggested action to take in connection with one of the discrete data elements of the data record. Output is generated representing the suggested action. A user provides input indicating acceptance or rejection of the suggested action. The suggested action may be performed in response to the user input.

1 . A computer-implemented method performed by at least one computer processor, the method comprising:
 (A) receiving data in a first data representation;   (B) extracting free-form text from the first data representation;   (C) identifying, based on the free-form text, a plurality of discrete data elements of a data structure;   (D) identifying, based on the free-form text, the plurality of suggested actions, wherein the plurality of suggested actions comprises, for each discrete data element D in the plurality of discrete data elements of a data structure, a corresponding suggested action S comprising: storing a value in the discrete data element D, removing a current value of the discrete data element D, or performing a specified operation on a current value of the discrete data element D to produce a new value of the discrete data element D; and   (E) performing the at least one of the plurality of suggested actions.   
     
     
         2 . The method of  claim 1 , wherein the first data representation comprises at least one of an audio signal representing speech and a transcript of the audio signal representing speech. 
     
     
         3 . The method of  claim 2 , wherein (B) comprises:
 (B1) detecting the format of the first data representation;   (B2) when an audio signal is detected, converting, using automatic speech recognition, the audio signal into a transcript corresponding to the speech present in the audio signal and parsing the transcript to extract the free-form text; and   (B3) when a transcript is detected, parsing the transcript to extract the free-form text.   
     
     
         4 . The method of  claim 1 , wherein the data structure comprises an electronic medical record. 
     
     
         5 . The method of  claim 1 , further comprising:
 (F) after identifying the plurality of suggested actions, providing output representing, for each discrete data element D in the plurality of discrete data elements of the data structure, the corresponding suggested action S.   
     
     
         6 . The method of  claim 5 , further comprising:
 (G) receiving a user input indicating user's acceptance of the at least one of the plurality of suggested actions, wherein the at least one of the plurality of suggested actions is performed in response to the receiving the user input.   
     
     
         7 . The method of  claim 6 , further comprising:
 (H) storing a record of the at least one of the plurality of suggested actions and the user input;   (I) repeating (A)-(G) at least once to produce a plurality of stored records of the plurality of suggested actions and a plurality of corresponding user inputs; and   (J) modifying a process used to identify the plurality of suggested actions based on the plurality of stored records.   
     
     
         8 . The method of  claim 7 , further comprising:
 (K) displaying output representing outcome of performing the at least one of the plurality of suggested actions.   
     
     
         9 . The method of  claim 1 , wherein (D) comprises:
 (D)(1) identifying a concept based on the free-form text;   (D)(2) determining, for each discrete data element D in the plurality of discrete data elements of the data structure, whether the concept is consistent with the discrete data element D; and   (D)(3) in response to determining that the concept is not consistent with the discrete data element D, identifying a suggested action, wherein the identified suggested action causes the discrete data element D to be consistent with the concept.   
     
     
         10 . The method of  claim 1 , further comprising:
 (F) before receiving the free-form text, receiving a patient identifier, and   (G) before identifying the plurality of discrete data elements of the data structure, identifying the data structure associated with the patient identifier.   
     
     
         11 . A non-transitory computer-readable medium comprising computer program instructions executable by at least one computer processor to perform a method, the method comprising:
 (A) receiving data in a first data representation;   (B) selecting free-form text from the first data representation;   (C) identifying, based on the free-form text, a plurality of discrete data elements of a data structure;   (D) identifying, based on the free-form text, the plurality of suggested actions, wherein the plurality of suggested actions comprises, for each discrete data element D in the plurality of discrete data elements of a data structure, a corresponding suggested action S comprising: storing a value in the discrete data element D, removing a current value of the discrete data element D, or performing a specified operation on a current value of the discrete data element D to produce a new value of the discrete data element D; and   (E) performing the at least one of the plurality of suggested actions.   
     
     
         12 . The non-transitory computer-readable medium of  claim 11 , wherein the first data representation comprises at least one of an audio signal representing speech and a transcript of the audio signal representing speech. 
     
     
         13 . The non-transitory computer-readable medium of  claim 12 , wherein (B) comprises:
 (B1) detecting the format of the first data representation;   (B2) when an audio signal is detected, converting, using automatic speech recognition, the audio signal into a transcript corresponding to the speech present in the audio signal and parsing the transcript to select the free-form text; and   (B3) when a transcript is detected, parsing the transcript to select the free-form text.   
     
     
         14 . The non-transitory computer-readable medium of  claim 11 , wherein the data structure comprises an electronic medical record. 
     
     
         15 . The non-transitory computer-readable medium of  claim 11 , wherein the method further comprises:
 (F) after identifying the plurality of suggested actions, providing output representing, for each discrete data element D in the plurality of discrete data elements of the data structure, the corresponding suggested action S.   
     
     
         16 . The non-transitory computer-readable medium of  claim 15 , where the method further comprises:
 (G) receiving a user input indicating user's acceptance of the at least one of the plurality of suggested actions, wherein the at least one of the plurality of suggested actions is performed in response to the receiving the user input.   
     
     
         17 . The non-transitory computer-readable medium of  claim 16 , where the method further comprises:
 (H) storing a record of the at least one of the plurality of suggested actions and the user input;   (I) repeating (A)-(G) at least once to produce a plurality of stored records of the plurality of suggested actions and a plurality of corresponding user inputs; and   (J) modifying a process used to identify the plurality of suggested actions based on the plurality of stored records.   
     
     
         18 . The non-transitory computer-readable medium of  claim 17 , where the method further comprises:
 (K) displaying output representing outcome of performing the at least one of the plurality of suggested actions.   
     
     
         19 . The non-transitory computer-readable medium of  claim 1 , wherein (D) comprises:
 (D)(1) identifying a concept based on the free-form text;   (D)(2) determining, for each discrete data element D in the plurality of discrete data elements of the data structure, whether the concept is consistent with the discrete data element D; and   (D)(3) in response to determining that the concept is not consistent with the discrete data element D, identifying a suggested action, wherein the identified suggested action causes the discrete data element D to be consistent with the concept.   
     
     
         20 . The non-transitory computer-readable medium of  claim 11 , wherein the method further comprises:
 (F) before receiving the free-form text, receiving a patient identifier, and   (G) before identifying the plurality of discrete data elements of the data structure, identifying the data structure associated with the patient identifier.