Every day I am seeing multiple doctors and people associated with the essentials getting affected due to the community spread of coronavirus through the people who are either scared to test or others who do not even have any symptoms associated with the virus. I wanted to thus build a system to make them feel safer by tracking the undiscovered positive coronavirus cases by boosting the amount of testing that is done each day.
What it does
Our system works on a machine-learning algorithm to find out the common features and symptoms of the people who have coronavirus through the data entered by the hospitals and then use it to find if a person has a high vulnerability of having a virus or not. If yes, we built a health-tracking device that is worn by these people, at self-quarantine at their homes while it measures their heart rate, respiratory rate, oxygen blood separation and skin temperature for 14 days. These are then tracked on our dashboard which is again monitored remotely by the doctors and if seem at high risk then would be shifted for intensive care. This way the workload of the doctors at the hospitals also reduces.
How we built it
We used multiple machine learning algorithms like SVM, logistic regression, Naive Bayes, neural networks to optimize the learning give out the most common features. The wearable health device uses multiple sensors like ECG, SpO2, RR sensors are used to measure the parameters and then this data is sent through IoT to the dashboard where it is monitored by the doctors.
Challenges we ran into
collecting the data of individual patients and their features and symptoms
Accomplishments that we're proud of
We have been able to make an algorithm and successfully predict if a person is vulnerable or not and connect it to the server. If we get some more data from the hospitals and with the help of government permission we will be able to further make an even better and dependable platform.
What we learned
What's next for Ingenious COVID Tracking
We further plan to partner with multiple hospitals, pharma shops and get reliable data from there. A further partnership with Aarogya Setu would help us get us data. Further to this, we plan to reduce the cost of the health tracking device and get the data in real-time so that it is acceptable by people and doctors.
adobe, heartratemonitor, machine-learning, numpy, python, react, respiratoryratemonitor