Longitudinal data quality assurance system

Patent metadata

Field	Value
Publication number	US-11315665-B2
Application number	US-201916356002-A
Country	US
Kind code	B2
Filing date	Mar 18, 2019
Priority date	Nov 21, 2018
Publication date	Apr 26, 2022
Grant date	Apr 26, 2022

Abstract

A longitudinal data quality assurance system is operable to receive a set of medical scans corresponding to a same first patient. A first chronologically ordered list of the set of medical scans is generated based on a corresponding first set of dates, where each of the corresponding first set of dates are extracted from a headers of the set of medical scans. Quality assurance data is generated for the first chronologically ordered list by performing at least one quality assurance function on at least one of the set of medical scans. A second chronologically ordered list that includes a first subset of the first set of medical scans is generated to rectify at least one continuity error of the first chronologically ordered list, indicated in the quality assurance data. The second chronologically ordered list is transmitted to a client device for display via a display device.

First claim

What is claimed is:

1. A longitudinal data quality assurance system, comprising:
at least one processor; and
a memory that stores operational instructions that, when executed by the at least one processor, cause the longitudinal data quality assurance system to:
utilize artificial intelligence to train a computer vision model operable to detect abnormalities in image data of medical scans;
utilize artificial intelligence to train an inference function operable to generate a predicted chronological ordering for a set of unordered lesion measurements;
receive, via a receiver, a first set of medical scans, wherein a first unique patient identifier of the first set of medical scans correspond to a same first patient;
generate a first chronologically ordered list of the first set of medical scans based on a corresponding first set of dates, wherein each of the corresponding first set of dates are extracted from a header of a corresponding one of the first set of medical scans;

generate first quality assurance data via artificial intelligence for the first chronologically ordered list indicating at least one chronological ordering error of the first chronologically ordered list by performing at least one quality assurance function on at least one of the first set of medical scans, wherein performing the at least one quality assurance function includes utilizing the computer vision model trained via artificial intelligence to detect an abnormality in image data of a subset of the first set of medical scans, wherein performing the at least one quality assurance function further includes utilizing artificial intelligence to perform a lesion measurement function on the subset of the first set of medical scans to generate a set of abnormality size measurements for the abnormality in the subset of the first set of medical scans, and wherein performing the at least one quality assurance function further includes performing the inference function upon the set of abnormality size measurements utilizing artificial intelligence to generate a predicted chronological ordering for the set of abnormality size measurements, and wherein the first quality assurance data indicates at least one chronological ordering error of the first chronologically ordered list based on the predicted chronological ordering for the set of abnormality size measurements being different from an ordering of the set of abnormality size measurements indicated by the first chronologically ordered list;
generate a second chronologically ordered list that includes the first set of medical scans in response to the first quality assurance data indicating the at least one chronological ordering error in the first chronologically ordered list, wherein the second chronologically ordered list is generated to rectify the at least one chronological ordering error based on utilizing the predicted chronological ordering for the set of abnormality size measurements, and wherein the second chronologically ordered list is different from the first chronologically ordered list; and
transmit, via a transmitter, the second chronologically ordered list to a client device for display via a display device.

2. The longitudinal data quality assurance system of claim 1 , wherein the operational instructions, when executed by the at least one processor, further cause the longitudinal data quality assurance system to:
receive, via the receiver, a second set of medical scans, wherein a second unique patient identifier of the second set of medical scans correspond to a same second patient;
generate a third chronologically ordered list of the second set of medical scans based on a second corresponding set of dates, wherein each of the second corresponding set of dates are extracted from a header of a corresponding one of the second set of medical scans;
generate second quality assurance data for the third chronologically ordered list by performing the at least one quality assurance function on at least one of the second set of medical scans.

3. The longitudinal data quality assurance system of claim 2 , wherein the operational instructions, when executed by the at least one processor, further cause the longitudinal data quality assurance system to:
receive a medical report corresponding to one of the second set of medical scans;
wherein performing the quality assurance function on at least one of the second set of medical scans includes extracting an incorrect date from the medical report, wherein performing the quality assurance function further includes determining the incorrect date does not match one of the corresponding second set of dates of the one of the second set of medical scans, and wherein the operational instructions, when executed by the at least one processor, further cause the longitudinal data quality assurance system to:
generate a corrected medical report by modifying the medical report to replace the incorrect date with the one of the corresponding second set of dates of the one of the second set of medical scans.

4. The longitudinal data quality assurance system of claim 2 , wherein the second quality assurance data indicates one of the second corresponding set of dates is incorrect, generate a fourth chronologically ordered list by changing a list position of one of the second set of medical scans corresponding to the one of the second corresponding set of dates.

5. The longitudinal data quality assurance system of claim 4 , wherein the operational instructions, when executed by the at least one processor, further cause the longitudinal data quality assurance system to:
receive at least one medical report corresponding to the one of the second set of medical scans;
wherein performing the quality assurance function includes extracting a correct date corresponding to the one of the second set of medical scans in the at least one medical report, wherein performing the quality assurance function further includes determining the correct date does not match the one of the second corresponding set of dates, and wherein the list position of the one of the second set of medical scans is changed to generated the fourth chronologically ordered list based on the correct date.

6. The longitudinal data quality assurance system of claim 1 , wherein the operational instructions, when executed by the at least one processor, further cause the longitudinal data quality assurance system to:
train the lesion measurement function by utilizing artificial intelligence.

7. The longitudinal data quality assurance system of claim 1 , further comprising:
determining diagnosis data for the first set of medical scans based on a corresponding set of medical reports, wherein generating the predicted chronological ordering for the set of abnormality size measurements is based on the diagnosis data.

8. The longitudinal data quality assurance system of claim 7 , wherein the predicted chronological ordering for the set of abnormality size measurements is generated to be monotonically decreasing over time based on the diagnosis data indicating at least one of: an improving condition or a positive final condition; and wherein the predicted chronological ordering for the set of abnormality size measurements is generated to be monotonically increasing over time based on the diagnosis data indicating at least one of: a worsening condition or a negative final condition.

9. The longitudinal data quality assurance system of claim 1 , wherein the first quality assurance data indicates the first chronologically ordered list is out of order based on determining the set of abnormality size measurements compares unfavorably to abnormality size change trend data, wherein the abnormality size change trend data is generated by ut

Assignees

• Enlitic Inc

Inventors

• Poblenz Eric C
• Yao Li
• Lui Keith
• Lyman Kevin

Classifications

• G06V40/171

  Local features and components; Facial parts (eye characteristics G06V40/18); Occluding parts, e.g. glasses; Geometrical relationships · CPC title

• G06V30/19173

  Classification techniques · CPC title

• G06V10/25

  Determination of region of interest [ROI] or a volume of interest [VOI] · CPC title

• G06V10/764

  using classification, e.g. of video objects · CPC title

• G06Q10/06315 Primary

  Needs-based resource requirements planning or analysis · CPC title