Summary

• Scan: BodyScan
• Duration: <60 seconds
• Data Returned: Body circumference and composition (body fat %), and health indicators such as waist-hip ratio and Type-2 Diabetes risk.
• Processing: On-Device
• Platforms: iOS, Android
• iOS submodule size: <10MB. 
• Android submodule Size: <10MB.
• Min. Requirements: iPhones: iOS 12.1 or above. Android:  Android 8 or above, 64-bit, and OpenGL 3.1 support.
• Requirements to Operate: Internet connection, Remote Assets, User Input

‍

What is BodyScan?

BodyScan is a scan technology where a front and side capture of a user's body is used to determine body circumference and body composition measurements.

BodyScan provides a process to capture a person's front and side image which is then used to calculate their anthropometric measurements. BodyScan includes inspection technology of the captured images to confirm that the user adheres to a minimum set of conditions, such as checking that the user is within frame and in the expected pose position, so that a high degree of accuracy can be maintained when using the technology.

‍

How it Works

• BodyScan is hardware accelerated to run on-device for a near real-time experience.
• Front and side images undergo proprietary computer vision and machine learning processing modules.
• Individuals receive composition, dimensioning, and health risk results, while partners can benefit from aggregated data points.

‍

Key Stats:

• Overall Circumference Accuracy: 97.5%
• Chest Circumference Accuracy: 98%
• Hip Circumference Accuracy: 97%
• Waist Circumference Accuracy: 98%
• Thigh Circumference Accuracy: 97%
• Weight Inference (avg): 96.6%
• Repeatability: 98%

‍

Data Outputs

User Input Data

• Height, Weight, Sex

Layer 1

Direct scan outputs

• Chest Circumference
• Waist Circumference
• Hip Circumference
• Thigh Circumference
• Fat-Free Mass

Layer 2

Derived data based on Layer 1 outputs

• Waist-Hip Ratio
• Waist-Height Ratio
• BMI
• Body Fat Percentage

Layer 3

Contextual data derived by comparing Layer 1 and 2 outputs and comparing them against public datasets or health studies.

Additional user input required to provide contextual data - Age, Ethnicity, Question Survey

• Obesity Risk
• Central Obesity Risk
• Type 2 Diabetes Risk