AI for Pneumonia Detection
The risk of pneumonia is enormous for many, especially in the nations where billions face energy poverty and rely on polluting forms of energy. “The WHO estimates that over 4 million premature deaths occur annually from household air pollution-related diseases including pneumonia” – WHO
Pneumonia, an AI-driven web app developed by Let’s Nurture, is an easy to use Pneumonia detector from Chest X-ray Images.
The Challenge
Build an AI to detect whether a victim is suffering from pneumonia or not just by looking at chest X-ray images.
A CNN (convolutional neural network) model trained from scratch to classify the presence of pneumonia from a collection of chest X-ray images data. We at Let’s Nurture constructed a convolutional neural network model to get features from a given chest X-ray image.DatasetThe original dataset consists of three main folders (training, testing, and validation folders) and two subfolders containing pneumonia (P) and normal (N) chest X-ray images, respectively. This dataset was organized and carried out as part of patients’ routine medical care. The dataset can be downloaded from the Kaggle website which can be found
here.
How does CNN actually see the images?
A CNN (convolutional neural network or ConvNet) is one of the most used and effective algorithms for deep learning, in which a model learns to perform classification tasks directly from images, video, text, or sound.
Image Kernels by Victor Powell
]In the past couple of years, cutting edge technology has started to become available to the software development community. Industrial strength packages such as Tensorflow have given us the same building blocks that Google uses to write deep learning applications for embedded/mobile devices to scalable clusters in the cloud and stochastic optimizers that make efficient applications possible.
Environment and tools
scikit-learn
keras
numpy
pandas
matplotlib
Results
After training, a separate test set was used to evaluate the accuracy of the CNN classifier. The trained model achieved an accuracy of 93%, which is substantially better than the baseline accuracy.AI-based Pneumonia detector might not be a marketable,/saleable product yet, but it makes me excited to see how easy it is to get started with. With even more open data we at Let’s Nurture are optimistic that we can achieve 100% accurate product.
Why choose Let’s Nurture for AI products?
We have been offering Python development services . We have a team of seasoned AI/ML engineers, who have been extensively working on healthcare domain for our customers.In this era of the technology revolution, Artificial Intelligence is strengthening its roots deep inside all verticals of telemedicine and remote healthcare industry practices If you want to know more about our skills and our ideas behind them, then why wait? Get in touch with us.
