Image analysis for predicting body weight in humans
AssigneeThe Children's Hospital of Philadelphia
InventorsJEFFREY S. BARRETT, CHEE NG
Life of this patent
Estimating the term…
Abstract
Systems and methods for determining body weight predictions and human conditions are disclosed. A body weight may be predicted by capturing at least one image of a human, and determining, from the image, a body weight prediction of the human by processing the at least one image with a data processor. The body weight prediction may further be based on an age-based weight factor. A model such as a neural network model may be used to predict body weight.
Claims19
A system for predicting body weight of a subject human, the system comprising:
- an image source configured to capture images;
- a data input device;
- a storage component configured to store an age-based weight prediction model and a trained neural network model, the trained neural network model trained with parameters corresponding to at least two facial feature measurement factors and a weight factor obtained from each of a plurality of other humans, the at least two facial feature measurement factors including a first facial feature measurement factor extending in a first direction and a second facial feature measurement factor extending in a second direction different from the first direction;
- a data processor coupled to the data input device, the image source, and the storage component, the data processor configured to capture an image of a face of the subject human with the image source;
- receive an age factor of the subject human with the data input device;
- determine the weight factor for the received age factor from the stored age-based weight prediction model;
- determine the at least two facial feature measurement factors for the face of the subject human from the captured image;
- and apply the weight factor and the at least two facial feature measurement factors to the trained neural network model to predict the weight of the subject human.
A method for predicting body weight of a subject human, the method comprising:
- capturing an image of a face of the subject human;
- receiving an age factor of the subject human;
- determining a weight factor for the received age factor from an age-based weight prediction model;
- determining at least two facial feature measurement factors for the face from the captured image, the at least two facial feature measurement factors including a first facial feature measurement factor extending in a first direction and a second facial feature measurement factor extending in a second direction different from the first direction;
- and
- applying the weight factor and the at least two facial feature measurement factors to a trained neural network model to predict the weight of the subject human, the trained neural network model trained with parameters corresponding to the at least two facial feature measurement factors and the weight factor obtained from each of a plurality of other humans.
A method of predicting the body weight of a human, the method comprising the steps of:
- capturing at least one image of the human;
- and
- determining, from the image, a body weight prediction of the human by processing the at least one image with a data processor.
A system for predicting the body weight of a human from at least one image of the human, the system comprising:
- a database, configured to receive and store a plurality of other human images and configured to receive and store a plurality of body weights associated with the plurality of other human images;
- and
- a data processor coupled to the database, configured to:
- receive the at least one image of the human from an image source;
- and
- determine by comparative analysis of the stored plurality of other human images and the stored plurality of body weights associated with the plurality of other human images in the database and the at least one received image, a body weight prediction of the human.
A system for determining a condition of a human based on a phenotypic response of the human from at least one image of the human, the system comprising:
- a database, configured to receive and store a plurality of other human images and configured to receive and store a plurality of phenotypic response data associated with the plurality of other human images;
- and
- a data processor coupled to the database, configured to:
- receive the at least one image of the human from an image source;
- and
- determine by comparative analysis of the stored plurality of other human images and the stored plurality of phenotypic response data associated with the plurality of other human images in the database and the at least one received image, the condition of the human.
Dependencies are inferred from the claim text.