You are on page 1of 2

Summary

Diabetes is a chronic disease which raises sugar level in blood and urine. It prevents human body
to energize itself from food energy. A healthy diet and physical activities are the best solution to
this disease. But in this age, the lifestyle of human beings has changed. Most people are
diagnosed with diabetes at critical stage where it becomes difficult to cure the disease. In this
research, a D-care (Diabetes Care) cloud-based monitoring solution has been given which will
continuously monitor the diet and physical activity of user to recommend him such a diet and
physical activity to balance the sugar level in blood/urine. D-care system will make use of
android and wearable fitness devices (IOT) to perform these functions. The internal structure of a
D-care system comprises of five integral parts which will be doing data collection, data
segregation, data interpretation and data storage along with providing the user with
recommended instructions. First part is called Front End. Its function is to detect and collect the
data based on human physical activity using the GPS and accelerometer sensor. However, the
user will feed his diet information manually. Second part is called Push API’s. It will retrieve the
collected data of all the sensing devices and logged by user on regular basis. It will further be
pushed towards the D-Care Engine or brain of the system. Third part is called D-care engine. It
will perform ETL(Extract-Transform-Load) and filter the data so that it can be processed without
any error. It will push the data to AWS (Amazon Web Services) Sage maker. AWS Sage Make
is the fourth part of D-care system. It works on different machine learning algorithms which
recommend the suitable diet and physical activity based on the analysis of received data. It filters
the received data to extract the relevant information for algorithm. The filtering process is
content based which filters the data based on past behaviors of user. It then generates two vectors
i.e. profile vector, containing the general user information, and item vector, containing the data
related to user’s diet and physical activity. It calculates the cosine value of both the vectors. Sage
Maker processes the cosine value to recommend exercise and diet to the user. This model is
based on the principles of machine learning and deep learning. It will keep on learning and
improving itself by analyzing the data over the course of time. The fifth and final part is called
D-Care Backend. It stores all the collected data and documents using Couchbase over the course
of time. This data can be accessed by the user to impact of recommendations on his health. This
data will keep on syncing with received data from AWS Sage Maker. This device syncs with the
daily life of the patient to learn about the activities of the patient. In this way, it will help not
only in early diagnosis of diabetes but also in curing the disease by recommending the patient a
healthy diet and required exercise.

You might also like