Page 248 - Big Data Analytics for Intelligent Healthcare Management

9.4 PROPOSED INTELLIGENT-BASED HRS 241

For example, when monitoring the health progress of patients and providing suggestions for treatment,
keep track of activities. In the case of food restriction for patients, the system has a difficult time mea-
suring its effectiveness, as some patients might eat without informing the system. Some health recom-
mendation may also aim at long-term behavioral changes and these must be tracked somehow, too.
Once the treatment is administrated, the system can continue monitoring the patient to determine if
treatment is effective. The system should also take steps to promote faster healing. We must consider
those recommendations that do not cause any side effect because neglecting one health parameter can
lead to another disease that is, changing food habits may lead to loss in body weight (a superficial health
parameter), keeping our body fit but neglecting a balanced diet can hamper growth and body metab-
olism. Before applying this approach in practical use, it must be ensured that the systems are customer
friendly and reliable. We must ensure that the system delivers real time results.

9.4 PROPOSED INTELLIGENT-BASED HRS
In this chapter, we put forward an intelligent-based HRS (shown in Fig. 9.5) supplied with big data
analytic tools to study and research health records of patients, assess risk and the severity of different
diseases, and then provide recommendations based on outcomes of prediction. This framework consists
of four main modules: the first part involves obtaining and accumulating the data from different sources
such as hospitals and medical centers, community health centers etc. Every patient has his own digital
record that includes demographics, patient prescription, previous medical history, lab tests, clinical test
results etc. Records are shared via secure centralized systems and are available for persons working in
healthcare centers. Every record is comprised of one updatable file, which means that doctors can carry
out changes and makes modifications with secured access. Since there are no duplicate files, there is no
need to worry about data replication. The second part of the framework refers to data preprocessing.
This part is responsible for processing and analyzing the huge quantities of collected data. Different
techniques such as attribute selection methods and data transformation methods are used to determine
and clean unnecessary data and find outliers. Insignificant and redundant attributes from data that do
not contribute to the accuracy of a predictive model are removed. This process is also known as data
cleaning. Data cleaning is essential to evade the formation of obscure or doubtful models and improve
the learning model performance. Furthermore, the data is also converted into a form appropriate for
classification. Therefore, data cleaning is a fundamentally important measure taken on raw data to pre-
pare it for the next stage.

Recommender
Data sources Data processing Data analysis engine Data visualization

Suitable medicine
Hospitals Data selection Big data Medical image Recommendations recommendations
Clinics Data tools(Hadoop diagnosis Health insurance
plan
transformation eco- Identification of
Patients records components) high-risk patients Health monitoring
Outlier analysis Usage patterns of of patients
Biometric scans drugs Medical reports
FIG. 9.5
Health recommendation system (HRS) framework.