Abstract:
"
Pneumonia is the leading cause of childhood death worldwide. In 2017 World Health
Organization published pneumonia killed more than 800000 children worldwide and children
under the Age of 5 are the very common victims of pnemonia.15 percent of all deaths are among
children are under the age of 5. According to one study a child dies every 20 seconds due to
pneumonia. Over 90 percent of pneumonia cases resulting in dead occur in developing countries.
In 2018 World health organization data published the latest data which shows influenza and
pneumonia deaths in Sri Lanka reached 4,864 or 3.83 percent of total deaths. The death rate is
21.48% per 100000 People. Pneumonia is mainly increasing deaths among children in
Sri Lanka.
A chest x ray is often used to diagnose pneumonia, but the x-ray results depend on the expertise
of the Radiographers and the accuracy to read an x-ray but Unfortunately sometimes things can
go wrong due to human errors made, this will lead to experience a worsening of the medical
condition. In some case you may need more dangerous complicated or expensive treatment as a
result. A person may even die as a result of illness that can have been treated or preventable.
Also, in reality this process is time consuming that needs a significant effort, if performed
manually.
Therefore, the aim of this research is to develop a pneumonia detector that will help doctors and
radiologist by predicting pneumonia and also reduce to the human errors made.
In order to train the machine learning algorithm to develop predictive data models Kaggle played
a major role in the systems data source."
