Pneumonia detection using CNN

Pneumonia detection using CNN

The ideal pnuemonia detection on the go.

Pneumonia detection using CNN

Pneumonia detection using CNN

The ideal pnuemonia detection on the go.

The problem Pneumonia detection using CNN solves

World's doctor to population ratio is very low. The workload on doctors is intensely immense, hence decreasing their own average life expectancy. 2.5 million People died from pneumonia in 2019. Almost a third of all victims were children younger than 5 years, it is the leading cause of death for children under 5. So there is a need of to diagnose pneumonia at an early stage, hence lowering doctor's burden and boosting the efficiency of illness detection. So I saw this issue with a different perspective that how i can introduce AI into medical field. My solution to this problem, which has claimed countless lives, owing to the time and cost of testing and treatment, is to detect the onset of pneumonia using chest X-rays of patients. A deep learning model is utilized to collect images of pneumonia and normal chest X-rays from Kaggel dataset. It is cost effective as it saves all the transportation charges and all the money that a patient spends on different types of lab tests. And you ever thought how much time and pain does it take for different lab tests required to diagnose, hence prolonging your diagnose and treatment. This code can make your diagnose much faster and painless. This code would keep in environment friendly diagnosis as it will reduce the carbon footprints by keeping all the data in digital format and removing the need of physical copy of the X-rays and all the scanned data is stored digitally which is easier to transfer.

Challenges I ran into

For this project I learned image processing, as this is the first time I am using images as a data. Converting image pixel into array was also a challenge for me. Knowledge about When to optimize and when to layer was most difficult thing in this project. Working with CNN also came up with many challenges.

Discussion