Create An Account
List of 3 Cases
List of 1
Accuracy of Neural Network
Pneumonia is a disease that affects the lungs and is caused by bacteria, viruses, or fungi. One of our teammate's four year old sister had pneumonia and could not go to school. After researching further into the problem, we found that pneumonia world’s leading cause of death among children under 5 years of age. In addition, it costs the medical industry billions of dollars and puts a strain on medical insurance.
What it does
Since web and mobile technology is very common, we created two programs that use X-ray images to notify the patient on whether they have pneumonia or not.
How we built it
There were two pieces of software developed, one for mobile technologies and one for a website. The mobile application was coded with React Native Expo, which is gaining popularity rapidly. Google Firebase was used to store the images and assess the data. In order to create this, we gathered a data set with x ray images of lungs from kaggle. We tested and trained the data, which allowed us to be able to input files and receive a prediction/diagnosis on our X-ray.
Challenges we ran into
Some challenges were figuring out how to use the neural nets and thoroughly understanding the advanced math behind them. Another challenge was linking the backend to the front end app interface.
Accomplishments that we're proud of
We were able to use the webcam and the phone camera to accurately identify whether there is pneumonia or not. Specifically for the mobile application, we were able to use login with Facebook, Google Plus, and Twitter. The neural network was able to detect a high level of accuracy (93%).
What we learned
We learned how to create neural networks and image recognition. In addition, we have learned how to integrate
What's next for Living Lungs
We want to improve the accuracy from 93% to 95%. We also want to expand to other systems and organs that can be tested through mobile applications.