Skip to main content

mhealth platform for Parkinson’s disease management

Periodic Reporting for period 1 - PD_manager (mhealth platform for Parkinson’s disease management)

Reporting period: 2015-01-01 to 2016-06-30

Parkinson’s disease (PD) is a complex chronic disease, individual disorder that most people live with for many years/decades. More than one million people live with PD in Europe today and this number is forecast to double by 2030. Also, there are predictions that by 2020 the number of diagnosed patients will be more than 12 million worldwide. It is the second most common neurodegenerative disease (after Alzheimer) and its prevalence will continue to grow as the population ages. Due to the disease complexity a multidisciplinary management involving several professions working together (neurologists, physiotherapists, speech and language therapists, occupational therapists, dieticians), is important to ensure that the patient retains his/her independence and continues to have the best quality of life possible. In the same context the role of the caregiver is of paramount importance.

The main objective of the PD_manager project is to build and evaluate an innovative mHealth ecosystem for PD management. PD_manager is a research project funded by the European Commission under the Horizon 2020 programme. Specifically, PD_manager was submitted in the PHC-26 topic Self-management of health and disease: citizen engagement and mHealth aimed at “Empowering citizens to manage their own health and disease will result in more cost-effective healthcare systems by improving utilisation of healthcare, enabling the management of chronic diseases outside institutions, improving health outcomes, and by encouraging healthy citizens to remain so.”

The PD_manager project aims to (see Figure 1):
• Use a set of mobile and wearable devices that will be used for symptoms monitoring and collection of adherence data.
• Assess motor and non-motor symptoms in Parkinson’s patients.
• Evaluate patients’ adherence to medical prescriptions.
• Conduct a dedicated nutritional study and empower game-based physiotherapy at home.
• Provide personalized suggestions for an optimal PD management plan.
• Propose an open architecture to support any commercial set of sensors within the Internet of Things concept.
• Model the behaviour of patients, caregivers, neurologists and other health-care providers.
• Educate patients, caregivers and healthcare providers with focus on occupational and speech therapies.

Different devices and wearable sensors will be used within PD_manager to carry out the continuous monitoring as well as to enable the performance of phone-based test and the delivery of education and training (see Figure 1):
• The wristband to continuously capture heart rate patterns, motion (with a 3-axis accelerometer) and skin temperature.
• The sensor insole, which is a novel product of Moticon ( is an every-day, flexible and thin solution to measure distribution of pressure, acceleration, weight-bearing, balance and motion sequences.
• A smart pillbox, which will be used to track the medication adherence of the patient.
• A smartphone, which will be used to capture the other elements of the mHealth platform as well as to deliver some specific tests, training and educational material. Caregivers and clinicians are also provided with smartphone solutions to participate in the care process.

The work within the project is organized into seven main work packages (WPs). WP1 (Management) deals with project management concentrated in handling the progress of actual work made within the project, its quality control as well as financial and communication issues. WP2 (Publicity and Business Potential) focuses on establishing communication channels, dissemination and publication of the project results, defining and updating the Data Management Plan and the Exploitation Plan as well as handling and resolving Intellectual Property Rights (IPR) issues. WP3 (Needs and Tasks Analysis & Decision Making Models) aims at developing decision making models based on observations of users’ decision making in situ, the computati
In the first half of the project – as described below – we have made much progress towards implementing the above specified goals, and towards the formalisation of our approaches. We have also disseminated our work widely, through academic talks and publications, and also through public-facing activities and working with journalists. As an overview summary, our efforts in the first half of the project have led to:
• 8 papers and conference abstracts being published describing our technical contributions and presented in various venues (for further details see subsection 2.4 Task 2.1).
• A project web site containing all the relevant information about the project, including papers and other dissemination materials (
• Twitter account was opened at the beginning of the project ( and is constantly active.
• Mentions of the project in various public-facing outlets, and at various keynote talks.

The project is progressing fluently, with smooth collaboration among partners, and we are able to implement and experiment with more sophisticated app development. The rest of this section describes in more detail the main results achieved.
Prototype apps for covering several aspects for data collection and management of PD have been implemented and evaluated. Those prototypes and the other achievements are described below, with further details given in the reports for the individual work packages.
• In WP3 needs and requirements from different users (prescribing clinicians, support clinicians, patients, caregivers, etc.) were elicited. Moreover, the models of clinicians’ diagnostic behaviour were obtained, which explain the clinicians’ decisions well, indicating that the information captured by PD_manager can effectively support their decision making. Furthermore, the state of the art of the currently used technologies in PD management were explored and based on this document the equipment for the experiments was ordered. An open architecture was designed to support the implementation of PD_manager system considering technological and user centric requirements. The architectural approach was also defined and was revised when GLOBO exited the consortium due to bankruptcy. Both the mobile and the cloud part are well defined and they are being implemented to provide the necessary infrastructure for the pilot study in 2017.
• In WP4 we have performed the first clinical study with PD patients. Based on this data, a detailed motor and non-motor symptoms analysis has been performed. Cognitive and speech mobile apps were developed. Moreover, based on a nutrition study, a mobile app was developed to track the nutrition habits and medication therapy of the patients. Also we have developed a gamification package for physiotherapy including the “Fruit Picking” game (for improvement of patient’s reach) and the “10 cubes” game (for improvement of fine motoric). The Educational gallery has been developed with 11 recorded videos that reflect some of the most common situations/symptoms of the patient suffering from PD. The development of the PD_manager Knowledge Management Platform is an undergoing task based on the principles of open architecture approaches and the Internet of Things concept, built in a combination and orchestration of cloud and mobile architectures, “blue-print” enterprise frameworks and open source solutions.
• In WP5 the data mining and decision support work are ongoing tasks. Several promising results were obtained: Initial models, built by using decision trees and neural networks for detecting ON/OFF stage of PD, have shown satisfying accuracy (90% for detecting ON, 60% for detecting OFF). Also, for the decision support task, a data analysis stage gave a number of models developed from the D3.5 dataset, where it turned out that DEX models were already better than the ones obtained with decision trees.
• In addition to the abov
PD_manager goes beyond the state of the art, since:

• We are using light and unobtrusive sensors for data capturing, especially the pressure sensors of the sensor insole haven’t been use in that context in the past.
• We are progressing well towards developing more accurate, specific and sensitive data analysis and decision support models based on past experience and background knowledge with the aim 1) to decide when medication change is needed (mainly based on the assessment of fluctuations and in the overall worsening of the patients) and 2) to automatically suggest an optimal medication plan that clinicians will approve.
• We are progressing in all parts of PD_manager’s holistic mhealth approach by combining several aspects for PD management: motor and non-motor symptoms detection, monitoring and evaluation, nutrition and gamification apps, educational gallery, decision support to manage the change of treatment, etc.

We have been particularly productive in terms of scientific dissemination of our work, with numerous papers published in top conferences and workshops. The feedback from the reviewers and the conference delegates has been very supportive, and we believe that our PD_manager system will become widely acceptable in the PD community. Moreover, our dissemination and Twitter activity has led to some interest from journalists at several national and international newspapers, who have reported on the PD_manager project. Taken together, all this gives us the confidence that the impact of the project will be substantial, and the usage of the apps will be widespread.

The potential impacts of the PD_manager project are listed below.

The informal caregiver (relatives, spouses, non-specialized nurses):
• Gets (daily or weekly) reports from the PD_manager nutrition app and the physiotherapy game in order to acknowledge the effort of the patient or to motivate-empower him/her to be more compliant. The caregiver also supports the patient to use the nutrition app and the game.
• Is sent a notification from the pillbox or the medication app for any missed pill in order to remind the patient take his medications.
• Is sent notifications when the patient needs to perform specific tests for cognition, speech, QoL, etc. and supports the patients do these tests.
• Is notified when the medication should be modified as indicated by the prescribing neurologist/ GP/ nurse through the PD_manager caregiver mobile app.
• Optionally, i.e. if the patient is instructed by the prescribing neurologist/ GP/ nurse, to wear all day long and carry with him a paired smartphone, the caregiver receives specific guidelines to support the patient. The caregiver activates activity and sleep monitoring modes in the band. The symptoms analysis is not exposed to the caregivers.
• Optionally, i.e. if the patient is instructed by the prescribing neurologist/ GP/ nurse, to wear insoles for gait assessment the caregiver again receives specific guidelines to support the patient. The gait analysis is not exposed to the caregiver.
• Is able to have sessions with psychologists/ psychiatrists through a secure video-therapy platform.
• Is educated about how to support the patient to cope with daily tasks and activities (occupational therapy) and manage speech, language and communication problems as well as eating, swallowing and drooling problems (speech and language therapy) and is more prepared for changes in the patient and increased symptomatology in order to stay in control of the disease management.

The patient:
• Uses daily the PD_manager nutrition app that includes nutrition and medication intake. He is also motivated to adhere to the nutrition plan.
• Does physiotherapy whenever instructed by the physiotherapist using the PD_manager game; the game on its own empowers the patient.
• Has the pillbox that provides data for the medication.
• Is sent notifications when he/ she needs to perform specific tests for cognition, sp
Monitoring equipment for the PD patient
Requirements for different users within PD_manager