IIT Delhi researchers have developed a web-based dashboard for predicting the spread of COVID-19 in India. The mobile-friendly dashboard, named as PRACRITI (PRediction and Assessment of CoRona Infections and Transmission in India) gives detailed state-wise and district-wise predictions of COVID-19 cases in India. The projections are given for a three week period, which is updated on a weekly basis.
A key parameter of interest on COVID-19 is the basic reproduction number R0 and its countrywide variability. R0 refers to the number of people to whom the disease spreads from a single infected person. For instance, if an active COVID-19 patient infects two uninfected persons, the R0 is two. Hence, reduction of R0 is the key in controlling and mitigating the COVID-19 in India.
Led by Prof. N. M. Anoop Krishnan, in collaboration with Prof. Hariprasad Kodamana, a team of volunteers from IIT Delhi, namely, Mr. Hargun Singh, Mr. Ravinder, Mr. Devansh Agrawal, Dr. Amreen Jan, Mr. Suresh, and Mr. Sourabh Singh have developed this dashboard.
PRACRITI provides the R0 values of each district and state in India based on the data available from sources such as the Ministry of Health and Family Welfare (MoHFW), Govt. of India; National Disaster Management Authority (NDMA), and World Health Organisation (WHO).
The predictions in the dashboard are based on a newly developed mathematical model that divides the population into four classes i.e. susceptible, exposed, infected, and removed. “Susceptible” refers to people who have not been exposed to the corona virus, “exposed” refers to those who have been exposed to the virus from an infected person, “infected” refers to those who are actively infected with COVID-19, and “removed” refers to those who are no longer a carrier of the virus.
Based on the computed values of R0, the researchers developed a detailed district-wise model for India to predict the number of actively infected person in each district. Further, to accommodate various effects due to administrative interventions, virulence of viral strain, change of weather patterns, the model will be updated on a weekly basis in an adaptive fashion to account these variations for accurate predictions.
-Shikhar Swami