Back to results
US 12,718,561 B2Granted
Protection of personally identifiable content in a video stream generated by an imaging device during a medical procedure
AssigneeIntuitive Surgical Operations, Inc.
InventorsYidan Qin, Maximilian H. Allan, Mahdi Azizian
Life of this patent
Estimating the term…
Abstract
An illustrative image processing system is configured to apply a video stream to a machine learning model, the video stream generated by an imaging device during a medical procedure performed with respect to a patient; classify, based on an output of the machine learning model, an image frame included in the video stream as an ex-body frame that depicts content external to a body of the patient: and apply, based on the classifying the image frame as the ex-body frame, a privacy enhancing operation to the image frame.
Claims18
A system comprising:
- a memory storing instructions;
- and
- a processor communicatively coupled to the memory and configured to execute the instructions to:
- apply a video stream to a machine learning model, the video stream generated by an imaging device during a medical procedure performed with respect to a patient;
- classify, based on an output of the machine learning model, an image frame included in the video stream as an ex-body frame that depicts content external to a body of the patient;
- apply, based on the classifying the image frame as the ex-body frame, a privacy enhancing operation to the image frame, the applying the privacy enhancing operation to the image frame comprising:
- determining a degree of blurring to be applied to the image frame based on a confidence interval of the output of the machine learning model, the confidence interval indicating a likelihood that the classifying of the image frame is correct and the degree of blurring representing a level of blurring to be applied to the image frame, and
- applying a blurring operation to the image frame in accordance with the degree of blurring;
- determine, based on the output of the machine learning model, that an additional image frame included in the video stream depicts content internal to the body of the patient;
- and
- abstain, based on the determining that the additional image frame depicts content internal to the body of the patient, from applying the privacy enhancing operation to the additional image frame.
8 more elements
A system comprising:
- a memory storing instructions;
- and
- a processor communicatively coupled to the memory and configured to execute the instructions to:
- apply a video stream to a machine learning model, the video stream generated by an imaging device during a medical procedure performed with respect to a patient;
- classify, based on an output of the machine learning model, a first plurality of image frames included in the video stream as in-body frames that depict content internal to a body of the patient;
- apply, based on the classifying the first plurality of image frames as in-body frames, a privacy enhancing operation to a second plurality of image frames included in the video stream, the second plurality of image frames not included in the first plurality of image frames, the applying the privacy enhancing operation to the second plurality of image frames comprising:
- determining a degree of blurring to be applied to the second plurality of image frames based on a confidence interval of the output of the machine learning model, the confidence interval indicating a likelihood that the classifying of the image frame is correct and the degree of blurring representing a level of blurring to be applied to the image frame, and
- applying a blurring operation to the second plurality of image frames in accordance with the degree of blurring;
- and
- abstain from applying the privacy enhancing operation to the first plurality of image frames.
7 more elements
A method comprising:
- applying, by an image processing system, a video stream to a machine learning model, the video stream generated by an imaging device during a medical procedure performed with respect to a patient;
- classifying, by the image processing system based on an output of the machine learning model, an image frame included in the video stream as an ex-body frame that depicts content external to a body of the patient;
- applying, by the image processing system based on the classifying the image frame as the ex-body frame, a privacy enhancing operation to the image frame, the applying the privacy enhancing operation to the image frame comprising:
- determining a degree of blurring to be applied to the image frame based on a confidence interval of the output of the machine learning model, the confidence interval indicating a likelihood that the classifying of the image frame is correct and the degree of blurring representing a level of blurring to be applied to the image frame, and
- applying a blurring operation to the image frame in accordance with the degree of blurring;
- determining, by the image processing system based on the output of the machine learning model, that an additional image frame included in the video stream depicts content internal to the body of the patient;
- and
- abstaining, by the image processing system based on the determining that the additional image frame depicts content internal to the body of the patient, from applying the privacy enhancing operation to the additional image frame.
5 more elements
Dependencies are inferred from the claim text.