System and method for correlation of pathology reports and radiology reports

Patent metadata

Field	Value
Publication number	US-10901978-B2
Application number	US-201415035772-A
Country	US
Kind code	B2
Filing date	Nov 14, 2014
Priority date	Nov 26, 2013
Publication date	Jan 26, 2021
Grant date	Jan 26, 2021

Abstract

A system for correlating patient radiology and pathology reports to track discordance among radiology and pathology diagnoses includes a natural language processor engine which extracts radiological information and pathology information. A correlation module correlates the radiology information and pathology information in a specific time period. A visualization graphical user interface indicates the correlation of radiology information and pathology information in a patient history. A tracking module which tracks misdiagnosis cases.

First claim

The invention claimed is:

1. A system for correlating patient radiology and pathology reports to track discordance among radiology and pathology diagnoses, the system comprising instructions that, in response to execution of the instructions by one or more processors, cause the one or more processors to:
perform natural language processing on clinical documents to extract radiology information and pathology information by identifying one or more predefined radiology study terms in radiology documents and one or more predefined pathology report terms in pathology documents, wherein the natural language processing extracts the radiology information and the pathology information according to the identification of the one or more predefined radiology study terms and the one or more predefined pathology report terms;
correlate the radiology documents and pathology documents in a specific time period by identifying matching anatomy according to a shared clinical concept, the radiology documents and the pathology documents respectively including the extracted radiology information and the extracted pathology information;
determine one or more misdiagnosis cases within the correlated radiology and pathology documents based on the extracted radiology and pathology information by identifying a mismatch of radiology diagnosis results with pathology results;
based on the correlated radiology and pathology documents and the extracted radiology and pathology information, identify the one or more misdiagnosis cases to calculate key performance indicators for use in a visualization graphical user interface;
generate data indicative of the visualization graphical user interface that indicates the correlation of radiology information and pathology information, wherein the visualization graphical user interface includes a visual scorecard that presents the calculated key performance indicators in spatial relation to accepted ranges of the key performance indicators; and
based on the data, render, on a display, the visualization graphical user interface.

2. The system according to claim 1 , wherein the natural language processing includes at least one of:
determining whether a current radiology report suggests any follow-up procedures; and
comparing a pathology report with a dictionary of known pathology result terms to find malignant diagnosis in the pathology report.

3. The system according to claim 1 , further comprising instructions to provide more information in response to determining a misdiagnosis case.

4. The system according to claim 1 , wherein one or more elements of the visual scorecard are selectable by a user to track more detail about each misdiagnosis.

5. The system according to claim 1 , further comprising instructions to track cases of misdiagnosis using a data structure to store information of correlation allowing for tracking of true positive diagnosis and false positive diagnoses.

6. The system according to claim 5 , wherein misdiagnosis information is tracked for a single radiologist over a specific period.

7. The system according to claim 5 , wherein misdiagnosis information for a department of radiologists at a hospital is tracked over a specified period.

8. The system of claim 1 , further comprising instructions to display to a user the correlations among radiology and pathology reports in a workflow for a user review of a patient history.

9. The system of claim 1 , wherein the natural language processing includes:
extraction of key words from a free-text portion of a clinician report; and
input of the keywords to a classifier to classify the clinician report as either a pathology report or a radiology report.

10. A method for correlating patient radiology and pathology reports, the method implemented using one or more processors and comprising:
extracting radiology information and pathology information from at least one radiology report and at least one pathology report by identifying one or more predefined radiology study terms in radiology documents and one or more predefined pathology report terms in pathology documents according to the identification of the one or more predefined radiology study terms and the one or more predefined pathology report terms;
correlating the radiology documents and pathology documents in a specific time period by identifying matching anatomy according to a shared clinical concept, the radiology documents and the pathology documents respectively including the extracted radiology information and the extracted pathology information;
determining one or more misdiagnosis cases within the correlated radiology and pathology documents based on the extracted radiology and pathology information by identifying a mismatch of radiology diagnosis results with pathology results;
based on the correlated radiology and pathology documents and the extracted radiology and pathology information, identifying the one or more misdiagnosis cases to calculate key performance indicators for use in a visualization graphical user interface;
generate and store in memory data indicative of the visualization graphical user interface, wherein the visualization graphical user interface includes a visual scorecard that presents the calculated key performance indicators in spatial relation to accepted ranges of the key performance indicators; and
based on the data, render, on a display, the visualization graphical user interface.

11. The method according to claim 10 , further including:
providing more information in response to determining a misdiagnosis case.

12. The method according to claim 10 , wherein one or more elements of the visual scorecard are selectable by a user to track more detail about each misdiagnosis case.

13. The method according to claim 10 , wherein the tracking includes using a data structure to store information of correlation allowing for tracking of true positive diagnosis and false positive diagnoses.

14. The method according to claim 10 , further including:
displaying to a user correlations among radiology and pathology reports in a workflow for user review of a patient history.

15. The method of claim 10 , further comprising:
extracting key words from a free-text portion of a clinician report; and
inputting the keywords to a classifier to classify the clinician report as either a pathology report or a radiology report.

16. A system for correlating patient radiology and pathology reports, the system comprising:
one or more processor programmed to:
extract radiology information and pathology information from at least one radiology report and at least one pathology report by identifying one or more predefined radiology study terms in radiology documents and one or more predefined pathology report terms in pathology documents according to the identification of the one or more predefined radiology study terms and the one or more predefined pathology report terms;
the radiology documents and pathology documents in a specific time period by identifying matching anatomy according to a shared clinical concept, the radiology documents and the pathology documents respectively including the extracted radiology information and the extracted pathology information;
based on the correlated radiology and pathology documents and the extracted radiology and pathology information, identify the one or more misdiagnosis cases to calculate key performance indicators for use in a visualization graphical user interface;
generate and store in memory data indicative of the visualization graphical user interface, wherein the visualization graphical user interface incl

Assignees

• Koninklijke Philips Nv

Inventors

• Xu Ye
• Qian Yuechen

Classifications

• G16H10/60 Primary

  for patient-specific data, e.g. for electronic patient records · CPC title

• G16Z99/00

  Subject matter not provided for in other main groups of this subclass · CPC title

• G16H15/00

  ICT specially adapted for medical reports, e.g. generation or transmission thereof · CPC title

• G16H40/20

  for the management or administration of healthcare resources or facilities, e.g. managing hospital staff or surgery rooms · CPC title

• G06F16/2365 Primary

  Ensuring data consistency and integrity · CPC title