Inspiration
Quite a lot of focus on social distancing and safety measures. But do we fully understand virus, given that it is continuously evolving? Is there way to be preventive and detect virus even earlier than when strong symptoms show? Testing everyone is a possible solution, but is it financially viable and practical, given that we would need to keep testing huge mass of population.
Instead what if everyone became proactive, by simply monitoring their own health. What if we could learn from medical history during incubation period of infected people. So that we can be faster at detecting virus at very early stage. Not only would this lower strain on healthcare should countries consider returning to "normal" movement of people, but very possibly virus would spread less fast. As unsolved danger of COVID-19 is the spread during incubation period.
Apart from currently being a web dashboard for public awareness purposes, Covid19 Health Monitoring can also integrate with contact tracing apps, by returning matrix of symptoms importance. That way without the user's health data leaving the user's phone, can the user be notified how their current health condition compares to being a serious COVID-19 case. That is probability of different scenarios of number of days till reaching that stage. Hence not only can this user refrain from going to public spaces, but can contact health authorities with better medical data to present (currently simply calling-in is quite bit of guess work for medical staff in call centers).
What it does
At present Covid19 Health Monitoring is a web dashboard, displaying equally weighted (if person has more than 1 symptom, different symptoms share weight of 1 patient) symptoms per age category and gender. Additionally average time till onset of confirmation and hospitalisation are displayed per age category and gender. Lastly list of diseases for patients with chronic history are also listed.
How I built it
After applying data science to found CSV file (Internet search), I modified process so that every important parameter can be a modifiable variable.
At end simply outputted data to PlotlyJS graphs.
Challenges I ran into
Finding medical data (for obvious privacy concerns). File I used, only 1% of records included symptoms, gender, age and realistic dates for all stages (first symptoms, confirmed case, hospitalisation).
Accomplishments that I'm proud of
Improving my skill set, learning about COVID-19 and hopefully contributing to solve crisis.
What I learned
Lot of symptoms out there! More precisely media only mentions cough and fever. But COVID-19 is more dangerous and we need a holistic approach.
What's next for Covid19 Health Monitoring
Is pending web endpoint that returns matrix of symptom importance. Before being able to do so, more medical data would be required to ameliorate decision of software. But Covid19 Health Monitoring has been designed in way that any CSV file can easily be imported, as one can specify names of columns of interest (age, symptoms list, chronic disease history list, date hospitalised, date confirmed, date symptoms first showed), as well as keywords in these columns that need to be considered. That way medical staff can also play around with tool!
Once more data is included, dashboard will also display current output offering for multiple days before showing strong symptoms. That way evolution of symptoms would be accessible to all.
Depending on popularity, integrate CSV upload web endpoint.
Log in or sign up for Devpost to join the conversation.