Back to results
US 7,627,152 B2Granted
Image-based indicia obfuscation system and method
AssigneeGE Medical Systems Information Technologies, Inc.
InventorsPrakash Parayil Mathew, Yaseen Samara, Vijaykalyan Yeluri, Denny Wingchung Lau
Life of this patent
Estimating the term…
Abstract
A technique is provided for obfuscating or masking indicia viewable in a reconstructed image. Image data is acquired and stored and then accessed and the reconstructed image is analyzed to determine whether any indicia are viewable in the image that are in need of masking. Such indicia may include, for example, patient-identifying data. Other information may be left unmasked, such as orientation information, descriptive information, dates, and so forth. The technique permits automated recognition of indicia and altering of image data files so as to render certain sensitive information undecipherable in the processed reconstructed images.
Claims22
A computer-implemented method for processing digital images comprising:
- on a physical computing processor:
- analyzing image data to identify indicia apparent in an image reconstructed from the image data;
- identifying one or more region in which the indicia appear in the image, including text in horizontal and vertical orientations;
- comparing the indicia to a list of indicia to remain decipherable or to a list of indicia to be rendered undecipherable in the reconstructed image;
- and
- based upon the comparison, replacing image data for at least one region with replacement data to render indicia undecipherable in an image reconstructed from the image data.
3 more elements
A computer-implemented method for processing digital images comprising:
- on a physical computing processor:
- analyzing image data via optical character recognition to identify textual indicia apparent in an image reconstructed from the image data;
- identifying one or more region in which the indicia appear in the image, including text in horizontal and vertical orientations;
- comparing the identified textual indicia to a list of textual indicia to remain decipherable in the reconstructed image;
- and
- based upon the comparison, replacing image data for at least one region with replacement data to render indicia in the at least one region undecipherable in an image reconstructed from the image data, and wherein textual indicia to remain decipherable in the reconstructed image is not replaced with replacement data.
3 more elements
A computer-implemented method for processing digital images comprising:
- on a physical computing processor:
- analyzing medical diagnostic image data via optical character recognition to identify textual indicia including indicia of patient identity apparent in an image reconstructed from the image data;
- identifying one or more region in which the indicia appear in the image, including text in horizontal and vertical orientations;
- comparing the identified textual indicia to a list of textual indicia to be rendered undecipherable in the reconstructed image;
- and
- based upon the comparison, replacing image data for one or more regions in which textual indicia is to be rendered undecipherable with replacement data to render the indicia undecipherable in an image reconstructed from the image data.
3 more elements
A system for processing digital images comprising:
- means for analyzing image data to identify indicia apparent in an image reconstructed from the image data;
- means for identifying one or more region in which the indicia appear in the image, including text in horizontal and vertical orientations;
- means for comparing the indicia to a list of indicia to remain decipherable or to a list of indicia to be rendered undecipherable in the reconstructed image;
- and
- means for replacing image data for at least one region with replacement data to render indicia undecipherable in an image reconstructed from the image data based upon the comparison.
2 more elements
A system for processing digital images comprising:
- means for analyzing image data via optical character recognition to identify textual indicia apparent in an image reconstructed from the image data;
- means for identifying one or more region in which the indicia appear in the image, including text in horizontal and vertical orientations;
- means for comparing the identified textual indicia to a list of textual indicia to remain decipherable in the reconstructed image;
- and
- means for replacing image data for at least one region with replacement data to render indicia in the at least one region undecipherable in an image reconstructed from the image data, and wherein textual indicia to remain decipherable in the reconstructed image is not replaced with replacement data based on upon the comparison.
2 more elements
A system for processing digital images comprising:
- means for analyzing medical diagnostic image data via optical character recognition to identify textual indicia including indicia of patient identity apparent in an image reconstructed from the image data;
- means for identifying one or more region in which the indicia appear in the image, including text in horizontal and vertical orientations;
- means for comparing the identified textual indicia to a list of textual indicia to be rendered undecipherable in the reconstructed image;
- and
- means for replacing image data for one or more regions in which textual indicia is to be rendered undecipherable with replacement data to render the indicia undecipherable in an image reconstructed from the image data based upon the comparison.
2 more elements
A computer program for processing image data comprising:
- at least one computer readable medium;
- and
- code stored on the at least one computer readable medium encoding routines for analyzing image data to identify indicia apparent in an image reconstructed from the image data, identifying one or more region in which the indicia appear in the image, including text in horizontal and vertical orientations, comparing the indicia to a list of indicia to remain decipherable or to a list of indicia to be rendered undecipherable in the reconstructed image, and replacing image data for at least one region with replacement data to render indicia undecipherable in an image reconstructed from the image data based upon the comparison.
A computer program for processing image data comprising:
- at least one computer readable medium;
- and
- code stored on the at least one computer readable medium encoding routines for analyzing image data via optical character recognition to identify textual indicia apparent in an image reconstructed from the image data, identifying one or more region in which the indicia appear in the image, including text in horizontal and vertical orientations, comparing the identified textual indicia to a list of textual indicia to remain decipherable in the reconstructed image, and replacing image data for at least one region with replacement data to render indicia in the at least one region undecipherable in an image reconstructed from the image data, and wherein textual indicia to remain decipherable in the reconstructed image is not replaced with replacement data based upon the comparison.
A computer program for processing image data comprising:
- at least one computer readable medium;
- and
- code stored on the at least one computer readable medium encoding routines for analyzing medical diagnostic image data via optical character recognition to identify textual indicia including indicia of patient identity apparent in an image reconstructed from the image data, identifying one or more region in which the indicia appear in the image, including text in horizontal and vertical orientations, comparing the identified textual indicia to a list of textual indicia to be rendered undecipherable in the reconstructed image, and replacing image data for one or more regions in which textual indicia is to be rendered undecipherable with replacement data to render the indicia undecipherable in an image reconstructed from the image data based upon the comparison.
Dependencies are inferred from the claim text.