Our project is intended to assist in the identification of infected people, as well as facilitate the access to health services through a simple process using the WhatsApp application. Using technology we can answer to nearly every family and give each of them a suitable treatment, optimizing our health system as a hole. Also the data we can provide will feed further investigations that are simply impossible in the moment due to the lack of consistent data. With a better intelligence we can do a bigger difference.
Check the link below to speak with Dr. Li: (only pt-BR)
We got inspired to serve and guide people in this uncertain moment everyone is going through, offering our idea to stand for the world and help our people.
We needed to create something that was accessible and economically viable to the population. Through artificial intelligence we followed Brazilian families up, making trials through WHO metrics we clarified fake news, diagnosed and guided the necessary measures. We used ZIP code to build a database and map spikes in contamination, number of people per houses and risk groups, inputs to further studies.
We used NASA data which have satellite images and geography data to guide the bot to points where more people with contamination risks are more likely to be.
The chatbot was developed by Nodejs, hosted in a AWS server and its voice processor was programmed in Python, converting audios to spectrogram, training a neural net made by Tensorflow.
Hosted in IBM Watson to recognize pulmonary problems.
We had more difficulties in collecting data to realize screening process, pulmonary exams and formulate psychological questions, besides the social isolation forbid us to gather personally.
You can also check our ppt presentation for more details:
https://docs.google.com/presentation/d/11czbKFrlzhmF4fi_PQKaWFMzup20HfDyqUTMPwXQyXs/edit
Check out pitch elevator video:
We used the following data provided by:
https://sedac.ciesin.columbia.edu/mapping/popest/covid-19/
Used as demographic complementary data for risk zones based on cases and deaths for each 100.000 people, as well as the analysis of age, sex and urbanization distributions.
Used as complementary data for insights, identifying different levels of human activity and thus a projection of the virus behaviour. The idea is to look for possible hotspots of the pathogen.