• TIHM-Dataset
  • Summary of Data Records
    • Activity
    • Labels
    • Physiology
    • Sleep
    • Demographics
  • Running the code

TIHM-Dataset

TIHM: An open dataset for remote healthcare monitoring in dementia. The dataset is available on its corresponding Zenodo repository.

The data used in this study will be made available by the the Surrey and Borders Partnership NHS Trust. Please, contact research@sabp.nhs.uk and specify the purpose for accessing the dataset.

Summary of Data Records

The dataset is organised in five separate tables stored as separate CSV files, including, Activity, Sleep, Physiology, Labels and Demographics. Data can be cross-referenced across the files.

Activity

	Value Type	Number of Values	Description
patient_id	CategoricalDtype	56	hash code
location_name	CategoricalDtype	8	Hallway,Lounge,Fridge Door,Bedroom,Kitchen,etc.
date	dtype[datetime64]	N/A	from 2019-04-01 to 2019-06-30

Labels

	Value Type	Number of Values	Description
patient_id	CategoricalDtype	49	hash code
date	dtype[datetime64]	N/A	from 2019-04-04 to 2019-06-30
type	CategoricalDtype	6	Agitation,Body temperature,Weight,etc.

Physiology

	Value Type	Number of Values	Description
patient_id	CategoricalDtype	55	hash code
date	dtype[datetime64]	N/A	from 2019-04-01 to 2019-06-30
device_type	CategoricalDtype	8	Skin Temperature,Diastolic blood pressure,Heart rate,O/E - muscle mass,etc.
value	dtype[float64]	N/A	min: 0.0, max: 211.0
unit	CategoricalDtype	5	%,kg,mm[Hg],beats/min,etc.

Sleep

	Value Type	Number of Values	Description
patient_id	CategoricalDtype	17	hash code
date	dtype[datetime64]	N/A	from 2019-04-01 to 2019-06-30
state	CategoricalDtype	4	LIGHT,AWAKE,DEEP,REM
heart_rate	dtype[float64]	N/A	min: 37.0, max: 107.0
respiratory_rate	dtype[float64]	N/A	min: 8.0, max: 31.0
snoring	dtype[bool]	2	True or False

Demographics

	Value Type	Number of Values	Description
patient_id	CategoricalDtype	56	hash code
sex	CategoricalDtype	2	Male, Female
age	CategoricalDtype	3	(70, 80],(80, 90],(90, 110]

Running the code

We have provided raw data and guidelines on how to access, visualise, manipulate and predict health-related events within the dataset. The Jupyter Notebooks have been developed using Python 3.9. For reproducing the code, an Anaconda virtual environment is also included. The virtual environment can be created using the following line of code in the Anaconda Terminal:

conda env create -f tihm.yml

After creating and activating the virtual environment, each notebook can be run individually. Please be careful to change the DPATH variable in each notebook with the folder in which the dataset has been downloaded.

DPATH = '../Dataset/'