########################################################################
# Title:       GOTOV Human Physical Activity and Energy Expenditure Dataset on Older Individuals
# Creator:     E. Slagboom, M. Beekman, S. Paraschiakos, R. Cachucho, A. Knobbe
# Publisher:   Leiden University Medical Center (LUMC)
# Contributor: Leiden Institute of Advanced Computer Science (LIACS)
# Created:     March 2015
# Contact:     S.Paraschiakos (s.paraschiakos@lumc.nl), M.Beekman (m.beekman@lumc.nl)
########################################################################

This dataset contains logs of 35 older participants (>60 years old) performing 16 everyday activities while wearing different devices in several body locations. In detail:
- GENEActiv accelerometer: ankle, wrist, chest.
- Equivital accelerometer, physiological measurements: chest.
- COSMED K4b2, energy expenditure measurements: facemark, torso belt.
The activities included where: jumping, lying down, sitting on a chair, sitting on a sofa, sitting on a couch (feet on the couch), standing still, washing dishes, stacking selves, vacuum cleaning, walking stairs up, walking in slow, normal and fast pace, and cycling.

CITATIONs:
====================================================================================
[1]
Paraschiakos, S., Cachucho, R., Moed, M. et al. Activity recognition using wearable sensors for tracking the elderly. User Model User-Adap Inter (2020). https://doi.org/10.1007/s11257-020-09268-2

[2]
Stylianos Paraschiakos, Cláudio Rebelo de Sá, Jeremiah Okai, Eline P. Slagboom, Marian Beekman, and Arno Knobbe. RNNs on monitoring physical activity energy expenditure in older people, 2020. (https://arxiv.org/abs/2006.01169)

DOCUMENTATION:
====================================================================================
- GOTOV_Dataset_Details.pdf, details on data collection protocol, participants demographics and sensors used
- GOTOV_DataSummary.csv, csv file including dataset details per participant:
	- ID, participant's ID
	- trial_date, date and time when data collection started
	- gender 
	- age, in years
	- weight, in kg
	- height, in cm
	- bmi
	- gaAnkle, TRUE if data from GENEActiv on the ankle exist, FALSE otherwise
	- gaChest, TRUE if data from GENEActiv on the chest exist, FALSE otherwise
	- gaWrist, TRUE if data from GENEActiv on the wrist exist, FALSE otherwise
	- equivital, TRUE if data from Equivital exist, FALSE otherwise
	- cosmed, TRUE if data from COSMED exist, FALSE otherwise
	- Time per activity in seconds (17 columns)
	- Total activity time in seconds (sum of activities time)
- GOTOV_photos, photos during data collection
- A video example of the data collection can be found here: https://youtu.be/jvx5FGhqPxw
====================================================================================

SENSOR DATA:
====================================================================================
1. Folder -> Activity Measurements
====================================================================================
Folder per participant (35 in total), every folder includes data from GENEActiv (per ankle, wrist, chest) and Equivital (chest)
Every GENEActiv csv includes the columns:
- indexes*
- timestamp in UNIX format
- acceleration measurements on x, y, z  in g
- activity label (NA if is unknown)

Similarly, Equivital csv includes columns with:
- indexes*
- timestamp in UNIX format
- HR..bpm, heart rate in beats per minute
- BR..rpm, breath rate in breaths per minute
- Skin.Temperature..IR, infra-red temperature in Celsius
- Body.Position (All orientations, Includes fall detection and motion)
- Ambulation.Status
- Alert
- PWI, Physiological welfare index
- Device.Indications
- Belt.Off.alert
- Low.HR.Confidence
- HR.Confidence
- Low.BR.Confidence
- BR.Confidence
- Button.Press.Event
- ECG.Saturation
- Apnea
- Irregular.rhythm.HR
- Heart.rate.hight.low
- Breathing.rate.high.low
- ECG.Breathing.rate.high.low
- ECG.BR..rpm
- ECG.BR.Quality
- ECG.Lead.1
- ECG.Lead.2
- y, acceleration on y-axis in mg 
- x, acceleration on x-axis in mg
- z, acceleration on z-axis in mg
- Breathing.Wave
- Inter.Beat.Interval..ms in ms
- label, activity label (NA if is unknown)

*indexes that start from numbers i>1 means that the preceding data input are deleted in order to have every devices starting at the same timestamp.

====================================================================================
2. Folder -> Energy Expenditure Measurements
====================================================================================
CSV files per participant including data of COSMED merged with ankle, wrist and chest accelerometer data of GENEActiv. Furthermore the files also include activity labels (true and predicted with [1])
Every COSMED csv includes the columns:
- time, timestamp in UNIX format
- acceleration measurements on x, y, z  in g
- ANKLEset_row/ WRISTset_row/ CHESTset_row, the original indexes of GENEActiv ankle, wrist, chest data (used for merging)
- COSMEDset_row, the original indexes of COSMED data (used for merging)
- ankleSVM/ wristSVM/ chestSVM, Signal Vector Magnitude per second of GENEActiv ankle, wrist, chest accelerometer,
- EEm,    Energy Expenditure per minute, in Kcal
- EEh, 	  Energy Expenditure per hour, in Kcal
- EEtot,  Total Kcal spent (it is reseted between indoor and outdoor measurements)
- METS,   Metabolic Equivalent per minute
- Rf,     Respiratory Frequency (litre/min)
- BR,     Breath Rate
- VT,     Tidal Volume in litre
- VE,     Expiratory Minute Ventilation (litre/min)
- VO2,    Oxygen Uptake (ml/min)
- VCO2,   Carbon Dioxide production (ml/min)
- O2exp,  Volume of O2 expired (ml/min)
- CO2exp, Volume of CO2 expired (ml/min)
- FeO2,   Averaged expiratory concentration of O2 (%)
- FeCO2,  Averaged expiratory concentration of CO2 (%)
- FiO2,   Fraction of inspired O2 (%)
- FiCO2,  Fraction of inspired CO2 (%)
- VE.VO2, Ventilatory equivalent for O2
- VE.VCO2,Ventilatory equivalent for CO2
- R,      Respiratory Quotient
- Ti,     Duration of Inspiration (seconds)    
- Te,     Duration of Expiration (seconds)
- Ttot,   Duration of Total breathing cycle (seconds)
- VO2.HR, Oxygen pulse (ml/beat)
- HR,     Heart Rate
- Qt,     Cardiac output (litre)
- SV,     Stroke volume (litre/min)
- label,  True activity label as noted from study protocol, NA if is unknown
- label,  Predicted activity label by model from [1], NA if is unknown
====================================================================================
