Two Forecasting Techniques Combined With AI To Approximate Flu Activity
Influenza is extremely contagious and can easily spread as individuals move about and travel. This makes the activity of tracking and forecasting flu a huge challenge. Whilst CDC is engaged in constantly monitoring patient visits for flu-like infections in the U.S., this data is mostly lagged behind for up to 2 Weeks than the actual time. A novel study, headed by the Computational Health Informatics Program (CHIP) at Boston Children’s Hospital, unites about two forecasting techniques with machine learning (part of artificial intelligence) to approximate regional flu activity. Outcomes of this research can be accessed in the journal Nature Communications.
When the approach called ARGONet was employed to flu seasons from September 2014 to May 2017, it offered more precise predictions than the team’s previous high-performing estimating approach, ARGO, in over 75% of the states studied. This highlights that ARGONet offers the most precise estimates of influenza activity available to date, a week ahead of conventional healthcare-based reports, at the state level across the U.S.
On a similar note, the updated guidelines disclosed by the Infectious Diseases Society of America (IDSA) highlighted that the pregnant women and highly obese people are amongst those at high risk for tricky situations from flu, it includes death as well. The guidelines also revealed that these people should get tested and started with the antiviral treatment as promptly as possible if they are hospitalized with the flu symptoms. Outpatients diagnosed with the flu and at high risk for impediments should also be offered antiviral treatment at the earliest, remarks the seasonal influenza guidelines.
These guidelines are available in the journal Clinical Infectious Diseases. They suggest the use of highly accurate and newer molecular tests, which offer results in 15–60 Minutes as a substitute to rapid-influenza diagnostic tests (RIDTs). Reportedly, RIDTs offer quick outcomes but can be wrongly negative in no less than 30% of influenza outpatients.