The Health Monitoring System module

The Health Monitoring System (HMS) has been developed following final end user's suggestions (parents of children with DCP), collected by means of a semi-structured questionnaire.

The whole system was optimized to be used in a domestic environment and/or clinical settings from a technical point of view and from a usability point of view. To improve the usability FSL, developed a lighter and user-friendly interface, running on a tablet.

The cardiac sensors and the accelerometer are embedded into the device. The breathing sensor is externally connected to it. The device is embedded in a band applied on the chest (Figure 10). All processing algorithms run on the Personal Computer which is connected via Bluetooth to the device. The tablet, running the graphical user interface (GUI) , is connected via Wi-Fi to the Personal Computer.

  • The GUI (Figure 11) has been developed on the basis of final end-user's suggestions: the number of buttons and needed operations required to achieve a goal were significantly reduced. Numeric and alphabetic information were drastically reduced as well, in favour of iconic information. The interface shows the heart rate (HR), the respiratory rate (RR) and the trunk orientation with respect to the external environment. The system also sends alarms both to the tablet of caregiver and to the ABC communicator and shows the last sent alarm. The interface is composed by: Two buttons displaying a numeric value, which describes the monitored parameter (HR, RR), an icon that visually indicates the parameter variation within the physiological range and a bar icon describing the quality of the acquired signal;
  • One button displaying the tri-axial trunk orientation through a head icon rotation;
  • An area listing the sent alarms;
  • A column showing information about the current monitoring session.

The end user can click the buttons displaying the related parameter graph (HR, RR or trunk orientation) plotted with respect of different timing.

The usability was evaluated directly by final end users (parents of children with DCP and physiotherapists) using structured questionnaires.

The HMS acquires cardiac, respiratory and accelerometric signals by means of a sensorized chest band and an additional sensor connected to it. The cardiac signal is acquired through three surface electrodes, the accelerometric signals are acquired through a tri-axial accelerometer positioned on the sternum with the axes of accelerometer coinciding with the antero-posterior, latero-lateral and cranio-caudal axis. All these sensors are embedded into the chest band. The additional sensor is a thermocouple placed within the nasal orifice acquiring the respiratory signal. The HMS allows for continuous monitoring of such parameters during the sleep/rest of DCP children with severe disability.

For the purpose of increasing the usability of the system, the parameters monitored by the HMS were limited to respiratory rate (RR), heart rate (HR) and trunk orientation in the space. These three parameters, in fact, have been scored as the most useful by end users in the administered structured questionnaires. The usability was improved by reducing the numbers of external sensors to apply and the number of cables of the system as suggested by end users.


Figure 11. Health monitoring system interface running on the tablet.

In terms of technical development, FSL designed the algorithms aimed at producing the above mentioned parameters. In a pre-processing phase (noise detection, baseline removal) all the recorded signals are calibrated and filtered. Final HR and RR parameters are both calculated by detecting heart peaks and breath peaks respectively in order to compare them with characteristic templates (of QRS complex and sinusoidal-like signals both determined in an initial baseline phase).  An orientation detection algorithm based on Euler angles calculation allows the detection of the trunk orientation with respect to the external space.

HR and RR values are continuously compared with predefined physiological thresholds. If such signal exceeds the relevant threshold, an alarm is sent to ABC communicator and to the tablet.

The three modules are dependent one to each other. During the whole processing chain, the accelerometer is used for artefact movement rejection, which prevents from unreliable processing of noisy hearth and respiratory signals.

The thresholds for alarm management are set, for different ages, by reviewing the values assessed in previous studies.

The above-mentioned signals are processed online in order to display a feedback to the end user on the interface. According to this approach, every signal is processed within a limited time window chosen as a trade-off between performance of the whole system and the specific biological signal analysis. Such algorithms were tested on healthy adult subjects, children with typical development and children with DCP.

Furthermore a new interface, running on a tablet, has been developed on the basis of final end-user’s suggestions. In general the whole system was optimized to be used in a domestic environment and validated in a real context.

On the other hand, FSL has also developed a separate system: the Physical Activity Monitoring (PAM). The PAM is intended to continuously monitor the quantity of movement performed by specific body segments in children with DCP and moderate motor disability.


Figure 12. The plot shows  the filtered breath signal, peaks identification and quality check (quality range=0-1) obtained processing the respiratory raw signal acquired from the thermocouple sensor during the resting phase.