Figure - PMC

Skip to main content

An official website of the United States government

Here's how you know

Here's how you know

Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS
A lock ( ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

View full-text article in PMC

. 2022 Mar 1;24(3):e27934. doi: 10.2196/27934

Search in PMC
Search in PubMed
View in NLM Catalog
Add to search

©Bo Zhang, Kaiwen Deng, Jie Shen, Lingrui Cai, Bohdana Ratitch, Haoda Fu, Yuanfang Guan. Originally published in the Journal of Medical Internet Research (https://www.jmir.org), 01.03.2022.

This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on https://www.jmir.org/, as well as this copyright and license information must be included.

PMC Copyright notice

Overview of data collection and streaming for meal activity analysis. (A) Data comes from two cohorts: 17 participants in the discovery cohort with 1658.98 hours of data and 17 participants in the validation cohorts with 2169.27 hours of data. (B) Signals were collected by 2 sets of sensors, accelerometer and gyroscope, in Apple Watch and paired with iPhone. (C) Z score and centering normalization were conducted for each of the x, y, and z axes for each window. (D) The gyroscope and accelerometer provide continuous signals on the x, y, and z axes over time. There are 2 modes to record meal time. The general model recording starts and stops by tapping the button on the Apple Watch. Retrospective mode allows the participant to type in the rough meal time with the iPhone after having a meal. For each record, we took 5-minute windows with a moving step of 20 seconds. The windows with >2.5 minutes of mealtime will be labeled as eating activity. Windows belonging to 2 sessions are removed. (E) Two augmentation methods, quaternion rotation and scaling the signal magnitude, apply to each data window. (F) Deep learning network structure.