Overview / Usage

Advanced Tools For Healthcare and Diagnostics Picture this-- your patient is unwell and he/she needs to get a battery of tests done. He/She spends the next few days going to different labs and then few more days waiting for the results. Once again he/she is back to you for a diagnosis. Now, here is a tool that lets you predict all the diseases and gives the results instantly.It is a tool developed for the convenience of clinicians for detecting various types of diseases especially subclinical illness.This web application deals with the prediction of potential sickness using machine learning techniques.The dataset used for this procedure are experimental datas.

Demo App: https://athad.herokuapp.com/
GitHub Repo: https://github.com/ADTHAD/ATHAD

Methodology / Approach

We have done various research in this project. There are currently 3 working tools in the Web App which can be used for detecting heart disease, diabetes in females and liver disease. We have used machine learning algorithms like Support vector machine, K-nearest neighbors and Logistic Regression for the predictions. We have used cleveland dataset for the training of heart disease model and, after the training we get accuracy of 88.52%. For diabetes prediction we used pima indian diabetes dataset for the training of the model and the accuracy is 81%. And For liver disease prediction we used indian liver patients dataset for the training of model.
We are also started working pneumonia prediction using chest X-Ray images.

Technologies Used

Intel AI DevCloud
GitHub

Flask
Python
HTML,CSS,JavaScript
Heroku

Libraries
Pandas
Tensorflow
Numpy
Scikit-learn
Matplotlib
seaborn
scipy
gunicorn

Repository

https://github.com/ADTHAD/ATHAD

Other links

• Demo App

Collaborators