mHealth specification
  • Introduction
  • Directory structure
    • Overview of directory structure
  • File types
    • Quick view of supported file types
  • sensor Data
    • Filename convention
    • File format
    • Sensor types
    • Data types
  • event Data
    • Filename convention
    • File format
    • Event types
  • annotation Data
    • Filename convention
    • File format
  • feature Data
    • Filename convention
    • File format
Powered by GitBook
On this page
  • File types to store raw information
  • sensor data
  • event data
  • annotation data
  • File types for data analysis
  • feature file
  1. File types

Quick view of supported file types

A quick review of the supported file types in mhealth-storage compatible dataset

File types are identified in the suffix before file extension, separated by symbol `.`. Current format reserves identifiers for six file types: sensor, event, annotation, feature, class and log. You may extend with your own file type identifier.

  • xxx.xxxxxx.2016-01-01-08-00-00-000-[M/P]HHmm.sensor.csv

  • xxx.xxxxxx.2016-01-01-09-00-00-110-[M/P]HHmm.event.csv

  • xxx.xxxxxx.2016-01-01-09-00-00-110-[M/P]HHmm.annotation.csv

  • xxx.xxxxxx.2016-01-01-09-00-00-110-[M/P]HHmm.feature.csv

  • xxx.xxxxxx.2016-01-01-09-00-00-110-[M/P]HHmm.class.csv

  • xxx.xxxxxx.2016-01-01-09-00-00-110-[M/P]HHmm.log.csv

File types to store raw information

sensor data

Store raw sensor data in high or low sampling frequency.

User can add their own sensor type to define new senosr data file format.

Filename convention

DeviceType-DataType-VersionCode.DEVICEID-SENSORTYPE.YYYY-MM-DD-HH-mm-ss-SSS-[M/P]HHmm.sensor.csv(.gz)
ActigraphGT9X-AccelerometerCalibrated-0x0x0.TASXXXXXXX-ACCEL.2018-01-01-11-00-00-000-P0000.sensor.csv

File format

HEADER_TIME_STAMP,X,Y,Z,[GROUP_1],[GROUP_2],...
2015-01-05 15:23:00.244,2.456,3.123,4.321,2.331,P1,TASXXXX,...
2015-01-05 15:23:00.244,2.456,3.123,4.321,2.331,P2,TASXXXX,...
2015-01-05 15:23:00.244,2.456,3.123,4.321,2.331,P3,TASXXXX,...

event data

Store formatted event information (numerical or categorical) happens at a time point, such as EMA prompt questions and responses.

Filename Convention

DeviceType-EventType-VersionCode.DEVICEID-EVENTTYPE.YYYY-MM-DD-HH-mm-ss-SSS-[M/P]HHmm.event.csv
Nexus5-EMAResponses-0x0x0.0123456789-EMARESPONSES.2018-01-01-11-00-00-000-P0000.event.csv

File format

PhoneCall
HEADER_TIME_STAMP,DIRECTION,NUMBER,DURATION_IN_SECONDS
2015-01-05 15:23:00.244,Out,617-xxx-xxxx,300
EMAResponses
HEADER_TIME_STAMP,PROMPT_TIME,RESPONSE_TIME,QUESTION,ANSWER
2015-01-05 15:23:00.244,2015-01-05 15:22:00.244,2015-01-05 15:22:03.244,"Where are you?","Home"

annotation data

Store annotation information by user or by annotator. Annotation set may be from user definition or predefined annotation set.

Filename Convention

AnnotationSet.ANNOTATIONSETID-ANNOTATORID-VERSIONCODE.YYYY-MM-DD-HH-mm-ss-SSS-[M/P]HHmm.annotation.csv

File format

HEADER_TIME_STAMP,START_TIME,STOP_TIME,LABEL_NAME
2015-01-05 15:23:00.244,2015-01-05 15:20:00.244,2015-01-05 15:21:00.244,Walking

File types for data analysis

feature file

Store numerical or categorical feature data derived from sensor or event files.

Filename Convention

FeatureType.FEATUREID-VERSIONCODE.YYYY-MM-DD-HH-mm-ss-SSS-[M/P]HHmm.feature.csv[.gz]

File format

HEADER_TIME_STAMP,START_TIME,STOP_TIME,[FEATURE_NAME],...
2015-01-05 15:23:00.244,2015-01-05 15:20:00.244,2015-01-05 15:21:00.244,1.44,...
PreviousOverview of directory structureNextFilename convention

Last updated 6 years ago