Smartphones have become an indispensable utility of our everyday lives. We use them day by day to communicate, as well as, produce and consume media such as photos and videos. Statistics show that penetration of smartphones in the ages of 55 to 64 years has reached a percentage of 61.5% with an increasing momentum [Source: ]. What it is taken usually as granted, it is the fact that smartphones constitute small computers with numerous sensors built in that allow for a significant array of functionalities.
By using the accelerometer, a sensor that can measure the (non-gravitational) acceleration of the phone in the three dimensions of space, and the gyroscope, a sensor that uses Earth’s gravity to help determine orientation [Source: ], your smartphone can “understand” how you hold it and if you move it. For example, based on the data sensed by the accelerometer, the phone is able to rotate its screen in prortait or landscape mode. Data from these sensors are also used by fitness applications as a basis to calculate a person’s level of movement activity.
In i-PROGNOSIS, we will use accelerometer and gyroscope data, i.e., how the smartphone is handled, to detect early minor tremor that may relate to Parkinson’s.
No details are required here. By using the microphone, usually, placed in the bottom of the phone, the device is able to transform your voice into an electric signal that will be transferred through a call or be recorded and stored as a digital file.
In i-PROGNOSIS, we will use samples of your voice during calls to detect voice degradation, such hypohonia, that relates to Parkinson’s.
No details are required here also. Every new smartphone, nowadays, has two cameras, a rear one and a front-facing one. The latter is the culprit for a recent trend in photo capturing called “selfies”, i.e., photos of ourselves captured, usually, with the camera located on the front of the phone.
In i-PROGNOSIS, we will process photos taken with the front-facing camera so as to recognise your face and facial expression to reveal possible masked-face features that relate with Parkinson’s.
Every new smartphone has a touch screen, a kind of screen technology that can sense touch, i.e., where, how and for how long the screen was pressed. It is the basic medium of interaction with the phone’s functions. Pressing, swiping, pinching are amongst the most common forms of touch that can be “understood” by the device.
In i-PROGNOSIS, we will use touching patterns, e.g., how you touch the screen in order to write a text or to unlock the phone, to detect early minor tremor that relates to Parkinson’s.
Smartwatches and fitness bands are a latest trend in technology and more specifically, wearable technology. As smartphones, they also constitute small computers disguised as watches and bracelets that, amongst others, bear sensors that allow for measuring physical activity and health-related parameters. Wearable usage in the age group of 55-64 years is expected to reach almost 28% by 2019, from 13.5% in 2015, in the US [Source: ], reflecting an impressive growing momentum.
Smartwatches and bands also have built in an accelerometer and a gyroscope. These sensors function in the same way as in the case of the phone; however, the main difference is that you wear them for long periods of time allowing for richer data regarding physical activity, as opposed to the phone that you usually leave on a desk.
In i-PROGNOSIS, we will use these data to detect the early onset of motor sympotms related to Parkinson’s, such as minor tremor or slowness of movement – bradykinesia.
By making use of an optical heart rate sensor, usually located on the back of the smartwatch/band, which can calculate the number of times the heart beats per minute based on the blood flowing through the wrist, smartwatches/bands can estimate your heart rate. The technology allowing for that is called which relies on the absorption/reflection of different colors of light.
In i-PROGNOSIS, we will use heart rate data as a general marker of health condition, as well as, combined with other measurements, to evaluate how well you sleep. Sleep disorders are an early symptom of Parkinson’s.
By making use of a temperature sensor, the wearable can take measures of the temperature on the surface of your skin and relate them to physical activity or even when and how well you sleep.
In i-PROGNOSIS, we will use temperature measurements along with heart rate data and data from the accelerometer and gyroscope to evaluate well you sleep and detect the onset of sleep disorders that are associated with Parkinson’s.
In i-PROGNOSIS we will use the Microsoft® Band™ that has all the aforementioned sensors built in. Not all smartwatches / fitness bands are equipped with these features.
Internet of things (IoT) refers to a network of physical objects that are embedded with electronics, software, sensors and network connectivity which enable them to collect and exchange data. It is a recent techological trend, known mostly for applications in connected homes, e.g., a thermostat or a lock that you can control using your smartphone from far away.
The Mandometer® is a personal scale that connects to a smartphone. It describes in real-time the progression of a meal and it offers directions helping the user to control the size of their meals by normalising the way they are eating. Patent for the Mandometer® is held by .
In i-PROGNOSIS, we will use the Mandometer® to track the way you eat your main meals in order to detect any dietary disorders, such as obesity, that are associated with Parkinson’s.
This can be easily categorised as a wearable sensor. The smart belt will be a new product, developed from scratch during the project. The device is conceptualised as an array of miniaturised microphones capturing bowel sounds and disguised as a belt that will be as comfortable as possible to wear at home during leisure time and sleep. The smart belt will be connected and exhange data with your smartphone.
In i-PROGNOSIS, we will develop the smart belt in order to detect the onset of frequent constipation that is one of the earliest symptoms of Parkinson’s.
Serious games are usually electronic games that are designed for a primary purpose other than pure entertainment. A wide spectrum of serious games has been designed and part of it is dedicated to health improving activities, such as exercise. A key term here is gamification, i.e., the disguise of tasks and activities into game playing in order to encourage engagement.
A gamesuite will be designed that will include electronic games targeting physical activity (Exergames), diet improvement (Dietary games), emotional expression (Emogames), handwriting and voice enhancement (Handwriting and Voice games). The games will be developed so as to be playable in a number of platforms such as the TV and the computer. Personalisation will be a key part of the games, as our intention is for the game activities to be adopted to the user’s needs and engagement.
In i-PROGNOSIS, the game suite will be part of the interventions aiming at sustaining the quality of life of Parkinson’s patients.
The Microsoft® Kinect™ is a sensor that allows you to use your body, voice and gestures to control and play electronic games. It has an camera to capture color images, an infrared (IR) emitter and depth sensor to capture also the depth of the image, multiple microphones to capture sound and an accelerometer for the device to “know” its orientation.
In i-PROGNOSIS, the Microsoft® Kinect™ sensor will be used as a controller for the game suite, especially in the case of Exergames where body acivity will be the main interaction.
Sensed data alone provide little information. The important things lie hidden and processing / learning algorithms come to the rescue. Hidden trends in measurements, characteristics of the signals and more can come to light through algorithmic processing and provide information about changes. The latter combined with software applications and the various sensors will form, what we like to call, the i-PROGNOSIS ecosystem.
Voice is sound that is converted to an electric signal by the microphone. To extract feaures of this signal that relate to Parkinson’s disease voice degradation, a number of steps must follow. A first step is to clear the recorded signal from noise so as to have just the signal corresponding to voice. Then, a number of algorithms will be applied to extract features that relate, amongst others, to the frequency, volume, and silent pauses of the signal. The latter will then be, somehow, combined in order to reach a conclusion on voice quality.
In i-PROGNOSIS, a number of processing algorithms will be developed and, combined with existing ones, they will be working in the background, processing the different types of data captured by the various sensors.
Imagine different types of data, captured frequently during the day, from a person for a long period of time and multiply it by thousands of people. This is pretty much what the term “Big Data” refers to. To combine these data and extract meaningful conclusions, special algorithms are needed, the so called machine learning algorithms. Machine learning algorithms are computer algorithms that are designed to learn from and make predictions on data.
In i-PROGNOSIS, early Parkinson’s detection will rely on a set of machine learning algorithms which will learn from the sensed and processed data in order to estimate the risk of developing the disease.
Mobile applications are the type of software applications developed for mobile devices such as a smartphone, a tablet or a smartwatch. They consist of a front-end – the user interface (UI) – that includes all the buttons and menus with which the user can interact and a back-end that encapsulates all the background services and algorithms that perform the various actions. Usually, they are called just “apps”.
In i-PROGNOSIS, we plan to develop mobile applications for both early Parkinson’s detection and supportive interventions. In the case of the former, the app will be publicly available for users to download and offer their data in order for our machine learning algorithms to learn.
The term Cloud refers to the storage / databases, data processing and exchange that takes place via third-party data centres, i.e., facilities with arrays of computers around the world. i-PROGNOSIS data storage and operations will heavily depend on the Cloud. In particular, all data collected will be stored on the Cloud, while machine learning algorithms and certain analysis algorithms will also run on the Cloud. All i-PROGNOSIS applications will exchange information with the Cloud.
i-PROGNOSIS will use the Microsoft® Azure™ cloud platform which complies with all the regulations concerning privacy and security of personal data.