Complete clinical data with data from patient environment
The Application Experiment will help homecare patients (i.e. sleeping apnea) to better monitor the evolution their treatments and maximise their acceptance. Monitoring services will be proposed for patient, Linde technicians and external prescribers.
Advances in several subfields of information and communication technology (ICT), including social computing, sensor networks, the Internet of Things, and intelligent information processing, have given rise to “a world of data” with fast and pervasive analytics at scale. In recent years, there has been a growing interest in collecting and analysing data about several aspects of our life, especially for home care. This emerging context has implications for how home health care may need to change to meet future needs.
In the context of home care, the adaptation of medical treatment to the Activities of Daily Living (ADL) remains a huge challenge when the knowledge about user lifecycle covers several dimensions: past activities, behaviour, beliefs, values, occupation, experiences, attitudes, etc. many health care data exist but that almost no information on home health care is available.
|LINDE||Data Analytics and patient monitoring||Extend data analytics experience to consolidate more consistent patient services and propose more diversified services for patient treatment monitoring|
|LYON2||IoT and CPS integration in industry, service engineering and Big Data analytics||Extend the patient environment with new sensors, IoT platform, CPS based actions in medical equipment’s, data concentration and transmission.
Data analytics to extend patient profile and propose new monitoring services to maximise treatment acceptance.
The Linde Group is a world leading supplier of industrial, process and specialty gases. Linde is one of the most profitable engineering companies proposing products and services in nearly every industry in more than 100 countries. A success story that began with the separation of air.
Linde Homecare France, a subsidiary company for homecare business, is specialized in patients’ following‐up with different chronical pathologies (respiratory, infusion, nutritional, etc.). In a permanent purpose of well‐being and respect of treatments compliance, the company is totally engaged in a e‐health strategy supported by connected devices and a new clinical approach for patients.
With the support of Lyon 2 research partner, the ongoing research activities target:
- Data analytics of clinical data,
- Patient profiling,
- Services adaptation to optimize health treatment and patient treatment path.
The following figure illustrates the ongoing data treatment flow:
The exploitation of collected data presents some limitations as the variability and the velocity of the patient data set is limited. Several patient parameters are missing and limit the relevance of data analytics efforts.
Problems to solve
To improve our patient related data set and consolidate a consistent profile, we identified from the literature the following missing parameters:
- Humidity level of the room
- The sleeping position of the patient
- Record the different stages of sleep
- Type of sleep apnea
- Detect when the mask is put on or not
- Cleanliness of the tube between the PPC (medical equipment) and the mask
- Outside temperature
- Room temperature (to deduce cooling or heating level)
- Parameters of the PPC (defined by the prescriber)
- Location where the PPC is used (which room?)
- Geo-localization of the PPC,
- Weight of the patient
- Heartrate of the patient
- Blood pressure of the patient
Our problem consists on identifying the needed distribution of sensors and actuators (to change the PPC parameters) and propose non-intrusive network at each patient premises with a reasonable cost and a high level of reliability.
The following figure presents the targeted scenario. We propose to complete the set of clinical data provided by the manufacturer of the medical facility with new data collected from the patient environment. We aim to propose new assistance services for the following stakeholders:
- The patient: to be informed about the performance of his treatment regarding the other patients in the same cluster.
- The prescriber: to be informed about the relevance of the PPC parameters.
- Linde: to be informed about the evolution of the patient acceptance of the treatment
Solutions to develop
With the support of ICT4Manuf DIH technologies, services and experimentations, Linde plan to develop the following business applications with the support of Lyon 2 partner:
- App1: Sensors network harvesting: generate a reliable and non-intrusive sensors’ network to collect the requested data for the missing parameters;
- App2: Real Time customer monitoring: mainly for Linde to detect and analyse interruptions in patient treatment
- App3: services for patients: to analyse the evolution of the treatment
- App4: services for prescribers: to validate the adequacy between patients’ profile evolution and the applied treatment.
|Apps||Requested services||DIH services|
|App1||Design network of sensors||ICT4Manuf(FR): Sensors’ network designer and IoT platforms|
|App2||Patient activity monitoring||ICT4Manuf(FR): Service designer for activity monitoring|
|App3||Bigdata analytics||ICT4Manuf(FR): Bigdata analytics & knowledge retention|
|App4||Services for events notification and monitoring||ICT4Manuf(FR): Service designer for external monitoring|
Target Key Performance Indicators
|Linde patients||5% increase in connected patients each year|
|Linde Operation management||100% of patient profiling for 2 therapies.
20% increase in patient acceptance of the treatment
|Linde prescriber partner||10% increase of the number of prescribers influenced by the proposed facilities.|