You are on page 1of 4

Support Health Monitoring with Cloud Business

Intelligence
André Sousa, Jorge Casaca, ISEP, Master in Computer Engineering

 temperature, and of the skin, among others, possible health risks


Abstract — This article intends to present a synthesis on the to patients monitored. For this we will develop a mobile
business intelligence in support system of health monitoring and application that will interact with the Power BI tool from
decision support for patients and doctors. In this context will be Microsoft.
presented a proposal solution for use data collected from patients
for their monitoring and counseling. Making use of different
existing technologies such as IoT devices and bracelets, we intend II. APP FOR HEALTH MONITORING
to present a methodology that makes use of Power BI and mobile It is intended to create a compatible real-time monitoring
platforms that allow doctors to accompany their patients. For this, system to use various portable sensors and user inputs to extract
the mobile application to be developed will have two strands: one
for the doctor and one for the patient. For the doctor, the
information useful for performing medical analysis and
application will monitor the patient and perform some actions, recommendations.
such as giving directions or triggering the means of emergency. The solution to be deployed must continuously power the
For the patient the application will be the entry point of the data cloud and thus leverage the capabilities of Microsoft Power BI.
for the Power BI and output with the doctor's notifications and The main features of this tool, which led to its choice, was
recommendations from the system itself. exporting reports, running R scripts and integrating with
Microsoft Azure. Figure 1 shows the architecture in our
Index Terms— Health Monitoring System, Decision Support
Systems, Business Intelligence, Mobile Health Monitoring, solution.
Microsoft Power BI, Cloud, IoT. The proposed system has the capacity to use several sensors
that allow simultaneous monitoring of various physical and

I. INTRODUCTION

W ITH the advent of communications the use of smart


mobile devices has received special attention from
researchers and entrepreneurs. Mobile devices such as
smartphones and tablets offer numerous opportunities to create
efficient mobile health solutions (m-Health).
The typical architectures of m-Health services use the
internet and web services to provide authentic interaction
between doctors and patients. A doctor or patient can easily
access the same medical record anytime, anywhere through
your personal computer, tablet or smartphone. The patient may
contact the physician in an emergency, or even have access to
medical records or appointments regardless of time and place.
The use of mobile devices by healthcare professionals has
transformed many aspects of clinical practice. Mobile devices
have become commonplace in healthcare settings, leading to
rapid growth in the development of medical software
applications for such platforms [3]. The various contributions
of m-Health focus on cardiology, diabetes, obesity, smoking
cessation, continuing care and chronic diseases [4].
Finally, we present the health monitoring solution supported
by Cloud BI. This solution is a real-time monitoring system to
use various portable sensors to extract information for
performing medical analyzes that help discover, through Fig. 1. The architecture solution for Support Health Monitoring with Cloud
multiple parameters, such as heart rate, blood pressure and body Business Intelligence.
2

chemical parameters of patients. For example, Zephyr BT [1] is III. MYHEALTH APP
an example of a sensor that captures patients' heart rates. Low The application developed is called "MyHealth". Regarding
power Bluetooth technology makes the sensor reliable to use the application layouts, they were intended to be simple for both
for a long period of time. Another example is blood pressure, the physician and the patient, for easy understanding and
which allows the Omron Wireless Upper Arm [2] to extract and usability of both parties.
transmit information to the cloud. This monitor is easy to use, For application use, doctors and patients must register. In
lightweight and has a long battery life. Figures 2 and 3 we have the login and registration screens,
Bracelets or other device that allows the extraction of data, respectively. The distinction between the doctor and the patient
such as heart rate, blood pressure, heart rate, respiratory rate, is fundamental to differentiate the operation of the application,
among others is done by means of sensors that these devices so much so that the application has a layout for the "doctor" and
have internally. We intend with this system to detect and predict another for the "patient" with different characteristics and
values outside the acceptable limits for the parameters functions. Validation of the user's features is the responsibility
monitored. Thus, if this happens, alarms will be triggered. of the application administrator.
These examples show the type of equipment that can be used
to extract information about various patient parameters for use
in the predictive models to be developed and integrated into the
proposed cloud system.
The amount of data extracted in real time for cloud storage
requires a well-planned architecture. From the data provided,
Power BI can provide physicians with useful information about
their patients and, from these reports, issue alerts or generate
recommendations for patients to change their behavior.
So far, we have talked about patients and doctors, however,
this is a more targeted solution. The design of this system
requires the definition of its target audience. Regarding patient
oriented mode there are no restrictions, so it can be used by an
8 to 80-years-old audience. However, the other mode of
application is more oriented to the so-called "family physicians" Fig. 2. Layout login. Fig. 3. Layout Registo.
because, given the type of data collected, they are the ones who
interact this type of information daily. In the following sections the two operating modes or
Technologically, the solution works on Microsoft Azure and application layouts are presented.
supports two modes of operation, as discussed above. For
patients, the application is exclusively mobile. For doctors the
application is web and has support for mobile. For the patient A. “Patient” Mode
and the sensors that it uses there will be a need to communicate In Figure 4 we can see the multilayer sketch for patient mode.
with Hub IoT. This is required to establish the communication The menus of the application's features are concentrated in the
of the bracelet or other device for communication with the left panel of the screen. On the right side, depending on the
cloud. The data received in the cloud is subsequently processed section chosen by the patient, the functions or functions of the
and distributed across the different data warehouses. Finally, patient are shown. In the upper right corner there is a
Power BI can analyze patient data and export results through notification area that can be accessed by clicking on the icon.
graphs, reports, tables, and more. The database used to store the These notifications can be alerts issued by the system,
data is SQL Server. recommendations, family physician messages or
By exporting analyzes performed by Power BI, doctors can tasks/activities that the patient has scheduled.
analyze the data and view them in graphs or another format. In the following sections we give a detailed description of
This support for doctors' decision-making will be made from each functionality of the existing application for the patient.
the office, home or other location where the physician can
perform the analysis and visualization of the data. This is Dashboard
possible because of the application's operation in the cloud. In the dashboard the user has an area with the panels that
Likewise, the appropriate functionalities will be made available he/she deems appropriate for his/her day-to-day life. By default,
to the patient. the application already comes with a dashboard mounted. These
If there are values outside the values delineated as normal, panels allow the information contained in the application to be
the alarms are generated indicating to the physician which displayed in reduced mode. For example, we talked about the
values are unusual. Likewise, alarms are triggered on the calendar/ schedule of activities and commitments; the message
patient's device. Along with the alarms, the application board exchanged with the doctor; and other panels with
supported by your artificial intelligence system will issue some historical charts of monitored physiological parameters. At the
recommendations to restore normality. All actions taken are top of the dashboard screen, the patient has the possibility to
properly archived. search through the search field.
3

B. “Doctor” Mode
In Figure 5 we can see the multilayer sketch for the medical
mode. The menus of the application's features are concentrated
in the left panel of the screen. On the right side, depending on
the section chosen by the patient, the functions or functions of
the patient are shown. In the upper right corner there is a
notification area that can be accessed by clicking on the icon.
These notifications can come from your own system, your
patients or contacts.
In the following paragraphs we give a detailed description of
each functionality of the existing application for the doctor.

Fig. 4. "Patient" Mode.

Monitoring
Here the user's monitoring capabilities are concentrated
through physiological parameters such as blood pressure
measurement, heart rate, oxygen, blood viscosity, respiratory
rate, ophthalmologic test, hearing test, lung capacity among
others. All parameters with temporal expression are represented
graphically. The application also provides tips to maintain or
reset the deviation of any parameter from normality. This is
done at the expense of the recommendation system developed
and implemented in the cloud. As a rule, no processing on the Fig. 5. “Doctor” Mode.
patient side of the application is performed. All analyzes are
performed on the platform server hosted on Azure. This allows Dashboard
for a more flexible and less costly developmental patient In the dashboard the doctor has an area with the panels that
application. he/she deems appropriate for his/her day-to-day life. By default,
Requests the application already comes with a dashboard mounted. These
The user has the possibility to interact with his doctor, asking panels allow the information contained in the application to be
for advice or scheduling consultations. It is also in this section displayed in reduced mode. At the top of the dashboard screen,
that the history of doctor-patient conversations is recorded. The the doctor has the possibility to search for what you want
idea of this section is not to be a patient-doctor chat. If the through the search field. This research tool is a working tool
patient wishes to interact with his doctor he must do it from this because it uses text-mining techniques to provide adequate
section. From here the patient can ask for appointments, response to the doctor's requests.
prescriptions, advice and view the recommendations issued by Patients
his doctor. The doctor has access to the patient records that accompanies
Schedule them. It can analyze all parameters monitored, the history of
It is a calendar and schedule of all the commitments and exams, consultations, prescriptions and conversations. You can
activities that the user has to perform. For example, we talk review the public part of your patient's schedule and add
about the time to take pills, the time to measure certain activities. You have the possibility to initiate a decision-making
physiological parameters, physical activity or the time of procedure regarding any fact about your patient. In this
consultation with your doctor. The doctor also has the procedure you can add doctors from your contacts to ask for an
possibility to schedule activities for your patient. opinion or help you in making any decision. All decision-
Configurations making procedures are properly archived and used by the
Finally, the settings section is for setting up the user account application's artificial intelligence system to generate
and managing personal data. For example, you may want to recommendations for further decision-making. These
change your password or change your email or account name. recommendations imply clustering techniques for patient
4

segmentation. Finally, this article aimed to make a state of the art on mobile
Reports devices for monitoring patients health, using mobile devices by
The Reports section allows your physician to review all the doctors and other health professionals, using Cloud BI
reports you have produced with Power BI. All analyzes and solutions, creating forecasting models using time series from
conversations, reports that the doctor produces are added to this data extracted from various sensors and mobile devices,
section and then attached to the patient records. coupled to users, both in the hospital environment and outside.
The use of Power BI together with support for the R language It was also intended to present a solution that uses several
allows the exploration of data science techniques that the technologies to monitor the health of patients using a Cloud BI
physician can use in its analyzes. solution.
Connecting this DSS to Power BI is intended to use the
capabilities of this tool in analyzing and generating reports for REFERENCES
structured data. From the data from the monitoring, the doctor [1] Medtronic; Available:
can perform predictive analyzes and thus anticipate eventual https://www.zephyranywhere.com/resources/hxm
complications for the patient. You can also use other Power BI [2] 10 Series Wireless Upper Arm Blood Pressure Monitor;
R support capabilities to analyze medical tests such as X-rays, Available: https://omronhealthcare.com/products/10-series-
TAC's, or other diagnostic exams. You can use clustering wireless-upper-arm-blood-pressure-monitor-bp786n/
[3] Mobile Devices and Apps for Health Care Professionals: Uses
techniques to classify your users to create better diagnostic and Benefits; Available:
models and prescribe better treatments. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4029126/
Schedule [4] Mobile-health: A review of current state in 2015; Available:
As in "patient" mode, the Schedule is available for the http://www.sciencedirect.com/science/article/pii/S15320464150
physician to manage your day-to-day, that is, your 01136
appointments. For example, appointments marked with
patients, analyzes of parameters of the various patients, reports
on the health status of patients, or other activities.
Wiki
The Wiki is a section the doctor has available to produce or
organize information about your daily activity. It's an extremely
popular Web format for creating shared documents. Generally,
there is no central wiki editor, no person who has final editorial
control. The purpose of this section is to create a database that
feeds the recommendation system referred to in the dashboard.
Contacts
It is used for the doctor to add the contact with whom it
interacts daily. These contacts can be used for addition to a
decision-making process that the physician initiates.
Configurations
Lastly, the settings section is the same as shown in the
"patient" mode.

IV. VII. CONCLUSION


In this article we explore the use of Business Intelligence
solutions hosted in the cloud to centralize the data, process it
and distribute it to interested parties. For this we develop a
mobile application that interact with the Power BI tool from
Microsoft. We intend with this solution to develop a service that
goes beyond existing applications that only perform static
analysis, that is, do not consider trends and seasonality.
The m-Health services and applications propose the delivery
of health care anytime, anywhere, overcoming geographical,
temporal, and even low-cost and affordable organizational
barriers [2].
The advent of mobile technology and applications is
transforming the way health information is accessed, delivered
and managed. Cloud computing is providing numerous benefits
to the healthcare industry by delivering and accelerating the
release of health services. The health industry's adherence to
cloud computing is inevitable and is already happening [2].

You might also like