Back to results
US 12,646,608 B2Granted
Automated visual reporting technique for medical imaging processing system
AssigneeBraid Health Inc.
InventorsKevin Quennesson, Alessandro Sabatelli
Life of this patent
Estimating the term…
Abstract
A medical imaging communication system generates an automated report for a medical image corresponding to a patient. The system receives one or more medical images for a patient. The system applies a model to identify an abnormality in the received one or more medical images. The model is configured to determine a likelihood score for the presence of the abnormality in the one or more medical images. If the likelihood score is above a threshold value, the system generates an automated report indicating the type of abnormality identified in the received one or more medical images. The automated report is then provided to a viewing user (e.g., a medical professional viewing the one or more medical images).
Claims18
A method comprising:
- receiving one or more images for a patient, the one or more images labeled with a body part captured in the one or more images;
- applying an image recognition model to determine a type of object present in the received one or more images;
- applying an abnormality detection model to determine a likelihood that the object present in the received one or more images is an abnormality based on the type of object determined by the image recognition model and the body part captured in the one or more images, the abnormality detection model trained to detect foreign objects not expected to be present at the body part captured in the one or more images using a training data set of images labeled with the type of object;
- and
- generating for presentation to a viewing user, responsive to determining that the likelihood is higher than a threshold value, an automated report indicating a type of abnormality identified in the received one or more images,
- determining a location corresponding to body part the received one or more images using a natural language processing model (NLP) trained to extract information from a medical report associated with the received one or more images, the medical reporting comprising text describing the location;
- and
- selecting the image recognition model based on the location corresponding to the body part.
5 more elements
A non-transitory computer readable medium comprising stored program code, the program code comprising instructions that when executed by one or more processors cause the one or more processors to:
- receive one or more images for a patient, wherein the one or more images are labeled with a body part captured in the one or more images;
- apply an image recognition model to determine a type of object present in the received one or more images;
- apply an abnormality detection model to determine a likelihood that the object present in the received one or more images is an abnormality based on the type of object determined by the image recognition model and the body part captured in the one or more images, the abnormality detection model trained to detect foreign objects not expected to be present at the body part captured in the one or more images using a training data set of images labeled with the type of object;
- and
- generate an automated report indicating a type of abnormality identified in the received one or more images upon a determination that the likelihood is higher than a threshold value,
- determine a location corresponding to the body part within the received one or more images using a natural language processing model (NLP) trained to extract information from a medical report associated with the received one or more images, the medical report comprising text describing the location;
- and
- select the image recognition model based on the location corresponding to the body part.
5 more elements
A system comprising:
- one or more processors;
- and
- a non-transitory computer readable medium comprising stored program code, the program code comprising instructions that when executed by the one or more processors cause the one or more processors to:
- receive one or more images for a patient, wherein the one or more images are labeled with a body part captured in the one or more images;
- apply an image recognition model to determine a type of object present in the received one or more images;
- apply an abnormality detection model to determine a likelihood that the object present in the received one or more images is an abnormality based on the type of object determined by the image recognition model and the body part captured in the one or more images, wherein the abnormality detection model is trained to detect foreign objects that are not expected to be present at the body part captured in the one or more images using a training data set of images labeled with the type of object;
- and
- generate, for presentation to a viewing user, an automated report indicating a type of abnormality identified in the received one or more images upon a determination that the likelihood is higher than a threshold value;
- determining a location corresponding to body part the received one or more images using a natural language processing model (NLP) trained to extract information from a medical report associated with the received one or more images, the medical reporting comprising text describing the location;
- and
- selecting the image recognition model based on the location corresponding to the body part.
8 more elements
Dependencies are inferred from the claim text.