# US 12,594,030 B1 · Regional body composition from two-dimensional color body images > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,594,030 B1 - Document ID: US12594030B1 - Status: Granted - Kind: B1 - Application number: 17/933,048 - Filed: 2022-09-16 - Granted: 2026-04-07 - Earliest claimed: 2022-09-16 - Assignees: Amazon Technologies, Inc. - Applicants: Amazon Technologies, Inc. - Inventors: Brandon Michael Smith; Siddhartha Chandra; Durga Venkata Kiran Yakkala; Ganesh Subramanian Iyer; Siddharth Choudhary; JinJin Li; Prakash Ramu; Antonio Criminisi; Steven Heymsfield - Web page: https://patentgrep.com/patent/US12594030B1 ## CPC - A61B 5/4872: {Body fat} - G06T 19/20: Editing of three-dimensional [3D] images, e.g. changing shapes or colours, aligning objects or positioning parts - A61B 5/7275: {Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor} - G06T 17/00: Three-dimensional [3D] modelling for computer graphics - G06T 15/005: {General purpose rendering architectures} - G06T 2207/10024: Color image - G06T 2207/20084: Artificial neural networks [ANN] - G06T 2207/30196: Human being; Person - A61B: DIAGNOSIS; SURGERY; IDENTIFICATION (usefulness limited to only animals A61D) - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL ## Abstract Described are systems and methods to determine one or more regional body composition values for one or more regions of a body based on a processing of one or more two-dimensional images that include a representation of the body. Regional body composition values may include any of a fat tissue mass, a lean muscle mass, a bone mineral density, a skeletal muscle mass, etc., which may be determined for any one or more regions of the body, such as the right arm, left arm, right leg, left leg, trunk, android, gynoid, etc. ## Claims ### Claim 1 A computer-implemented method, comprising: processing a two-dimensional (“2D”) color body image of a human body to determine a plurality of dimension based regional body composition values for at least one region of the human body; processing the 2D color body image to determine a plurality of image based regional body composition values for the at least one region of the human body; combining the plurality of dimension based regional body composition values and the plurality of image based regional body composition values to determine a plurality of regional body composition values for the at least one region; and causing a presentation of at least some of the plurality of regional body composition values determined for the at least one region of the human body. ### Claim 2 (depends on 1) The computer-implemented method of claim 1, wherein the at least one region includes at least one of a left arm, a right arm, a left leg, a right leg, a full body, a right bicep, a right forearm, a right upper leg, a right lower leg, a trunk, an android, or a gynoid. ### Claim 3 (depends on 1) The computer-implemented method of claim 1, wherein determining the plurality of regional body composition values includes: processing the 2D color body image to segment a first plurality of pixels of the 2D color body image that represent the human body from a second plurality of pixels of the 2D color body image that do not represent the human body to produce a silhouette of the human body; and processing the silhouette to produce the plurality of regional body composition values. ### Claim 4 (depends on 1) The computer-implemented method of claim 1, wherein: processing the 2D color body image to determine the plurality of image based regional body composition values, further includes: processing the 2D color body image to determine a plurality of personalized body features corresponding to the human body represented in the 2D color body image; and processing the plurality of personalized body features with a first neural network to determine the plurality of image based regional body composition values; and processing the 2D color body image to determine the plurality of dimension based regional body composition values, further includes: processing the plurality of personalized body features with a second neural network to determine a plurality of body dimensions of the human body; and processing the plurality of body dimensions with a third neural network to determine the plurality of dimension based regional body composition values. ### Claim 5 (depends on 1) The computer-implemented method of claim 1, wherein the plurality of regional body composition values include at least one of a fat tissue mass for the region, a lean tissue mass for the region, or a bone mineral density for the region. ### Claim 6 A computing system, comprising: one or more processors; a memory storing program instructions that, when executed by the one or more processors, cause the one or more processors to at least: receive at least one two-dimensional (“2D”) color body image of a body, wherein: each of the at least one 2D color body image is generated by a 2D imaging element; and each of the at least one 2D color body image includes a representation of the body from a different view; and process, using a neural network, each of the at least one 2D color body image to, at least: determine a plurality of dimension based regional body composition values for a region of the body: determine a plurality of image based regional body composition values for the region; and combine the plurality of dimension based regional body composition values and the plurality of image based regional body composition values to determine a plurality of regional body composition values for the region. ### Claim 7 (depends on 6) The computing system of claim 6, wherein the region of the body includes at least one of a full body region, a left arm region, a right arm region, a left leg region, a right leg region, a trunk region, an android region, or a gynoid region. ### Claim 8 (depends on 6) The computing system of claim 6, wherein the program instructions that, when executed by the one or more processors to determine the plurality of image based regional body composition values, further include instructions that cause the one or more processors to at least: process the at least one 2D color body image to determine a plurality of personalized body features corresponding to the body; and process the plurality of personalized body features to determine the plurality of image based regional body composition values. ### Claim 9 (depends on 6) The computing system of claim 6, wherein the program instructions that, when executed by the one or more processors to determine the plurality of dimension based regional body composition values, further include instructions that cause the one or more processors to at least: process the at least one 2D color body image to determine a plurality of body dimensions corresponding to the body; and process the plurality of body dimensions to determine the plurality of dimension based regional body composition values. ### Claim 10 (depends on 6) The computing system of claim 6, wherein; the program instructions that, when executed by the one or more processors to determine the plurality of image based regional body composition values, further include instructions that cause the one or more processors to at least: process the at least one 2D color body image to determine personalized body features corresponding to the body; and process the personalized body features to determine the plurality of image based regional body composition values; the program instructions that, when executed by the one or more processors to determine the plurality of image based regional body composition values, further include instructions that cause the one or more processors to at least: process the at least one 2D color body image to determine a plurality of body dimensions corresponding to the body; and process the plurality of body dimensions to determine the plurality of dimension based regional body composition values; and combine the plurality of image based regional body composition values and the plurality of dimension based regional body composition values to produce the plurality of regional body composition values. ### Claim 11 (depends on 6) The computing system of claim 6, wherein the program instructions that, when executed by the one or more processors to determine the plurality of regional body composition values, further include instructions that cause the one or more processors to at least: determine a weighting to apply to at least one of the plurality of image based regional body composition values or the plurality of dimension based regional body composition values; and combine the plurality of image based regional body composition values and the plurality of dimension based regional body composition values based at least in part on the weighting. ### Claim 12 (depends on 6) The computing system of claim 6, wherein the program instructions that, when executed by the one or more processors to process each of the at least one 2D color body image, further include instructions that cause the one or more processors to at least: combine, using the neural network, each of the plurality of dimension based regional body composition values, the plurality of image based regional body composition values and at least one body trait of the body to determine the plurality of regional body composition values. ### Claim 13 (depends on 12) The computing system of claim 12, wherein the at least one body trait is at least one of a height of the body, a weight of the body, an age, an ethnicity, or a gender of the body. ### Claim 14 (depends on 6) The computing system of claim 6, wherein the program instructions that, when executed by the one or more processors to cause the one or more processors to process each of the at least one 2D color body image, further include instructions that, when executed by the one or more processors, further cause the one or more processors to at least: process a first 2D color body image of the at least one 2D color body image to produce a first plurality of features representative of the body; and process a second 2D color body image of the at least one 2D color body image to produce a second plurality of features representative of the body. ### Claim 15 (depends on 14) The computing system of claim 14, wherein the program instructions that, when executed by the one or more processors to cause the one or more processors to process each of the at least one 2D color body image, further include instructions that, when executed by the one or more processors, further cause the one or more processors to at least: concatenate the first plurality of features and the second plurality of features to produce concatenated features; and process the concatenated features to produce the plurality of regional body composition values. ### Claim 16 (depends on 6) The computing system of claim 6, wherein the program instructions that, when executed by the one or more processors, further cause the one or more processors to at least: process, using a second neural network, each of the at least one 2D color body image to determine a three-dimensional model representative of the body; and send, for presentation, the three-dimensional model and at least one of the plurality of regional body composition values. ### Claim 17 A method, comprising: determining at least one region of a body represented in a two-dimensional (“2D”) color body image, wherein the at least one region is less than all of the body; processing the 2D color body image to determine a plurality of dimension based regional body composition values for the region; processing the 2D color body image to determine a plurality of image based regional body composition values for the region; and combining the plurality of dimension based regional body composition values and the plurality of image based regional body composition values to determine a plurality of regional body composition values corresponding to the at least one region; and sending for presentation the plurality of regional body composition values corresponding to the at least one region. ### Claim 18 (depends on 17) The method of claim 17, further comprising: generating a three-dimensional (“3D”) model of the body; and wherein sending includes sending for presentation the 3D model of the body and the plurality of regional body composition values. ### Claim 19 (depends on 17) The method of claim 17, wherein the plurality of regional body composition values include at least two of a left arm region, a right arm region, a left leg region, a right leg region, a trunk region, an android region, or a gynoid region. ### Claim 20 (depends on 17) The method of claim 17, wherein combining includes: combining the plurality of dimension based regional body composition values, the plurality of image based regional body composition values, and at least one trait about the body to determine the plurality of regional body composition values.