Sleep Monitoring and Assessment

Enora Innovation

Description

The ENORA Sleep Anaysis Toolkit is designed with PD clinicians and is focusing on the following services.

  • Position Clasification. This tool clasifies the grid data of the prototype ENORA pressure sensing mattress top to identify the specific category of PD status regarding the position of the patient on the bed.
  • Sleeping patterns analysis. This analysis receives as inputs a time series data and aims to detect known PD patterns. The position identification performed in Tool 1 is coupled with continuous data collected from the Withing sleep analysis sensor in order to reveal furtehr sleep insights. This tool can be used in combination with Tool 1 or independently exploiting the use of sleep analysis data specifically adapted to PD disorders.

This tool is combining:

  • Sleeping positions over night (Tool 1)
  • Continuous heart rate
  • Snore detection
  • Breathing patterns and breath disturbances
  • Sleep cycles
  • Sleep score

The assessment is based on the identification of specific PD disorders and classifies the PD patient medical condition with respect to PD sleep quality (with a level of uncertainty).

Documentation

Inputs and Outputs

As input we use:

  • ENORA mattress top pressure sensor data series (Tool 1)
  • Withings under mattress Sleep Sensor data series (Tool 2). Will require sensor authentication to use the data in ALAMEDA.
While as output:
  • PD sleep quality and medical condition assessment. This is work in progress with medical partners because no relevant standard exists

Background Info

Parkinson’s disease is a complex movement disorder that is affecting millions of people around the world. It’s estimated that two-thirds of those afflicted with Parkinson’s disease struggle to get quality sleep. In addition, sleep problems are increasingly recognized as a potential early indicator of Parkinson’s disease. ALAMEDA is directly addressing sleep quality for PD and the Sleep Monitoring & Assessment Toolkit for PD will provide assessment of the PD sleep quality and of their medical condition based on continuously collected data series for extended periods using non-intrusive/wear-nothing technologies. The use of a commercial device that is used Globally for sleep analysis will improve the outreach of ALAMEDA and will enable the use of the produced AI sleep analysis algorithms by users worldwide, via the ALAMEDA Innovation Hub.

Datasets & Samples

Currently the Sleep Monitoring Toolkit accepts JSON objects (key value pairs) of the sleep mattress and Withings sensor raw data. It returns position, out of bed, stillness. Data are pushed at 20 seconds interval (sleep mattress) and 60 seconds of the Withings (heart rate and respiration rate) This is a prototype version and the models are being upgraded with added modalities.

Installation Ιnstructions

The Sleep Monitoring Toolkit is packed in a dockerized container. Internally it starts a web browser listening to port 80. In order to run the container follow the steps below:

  • Login to container registry using guest account:

    docker login gitlab.telecom.ntua.gr:5050 -u alameda_ai_toolkit_registry_guest -p ByeYyNesUxqQGs91FzzW

  • Pull and run the docker image:

    docker run -d -p 80:80 gitlab.telecom.ntua.gr:5050/alameda/alameda-source-code/ai-toolkit/ai-toolkit-registry/coreapi

  • Logout from container registry:

    docker logout gitlab.telecom.ntua.gr:5050

OpenAPI Documentation

View Swagger Documentation