In what could be considered as a prime example of technology doing wonders, researchers from a university in the US have successfully developed a way through which the tweets posted on the 140 characters social networking giant could help in tracking the spread of seasonal flu, that too in real time.
The new model developed by the Northeastern University researchers utilizes Twitter tweets to predict how a particular infection might affect the human population on planet Earth. The researchers set out gathering tweets along with the parameters of each of the season’s epidemic, such as how many people an individual with the virus can infect, the incubation period of the disease, viral strains present and the immunisation rate etc.
Once the tweets were gathered, they progressed to the next step which involved applying forecasting and a couple of other algorithms to the key parameters of the infection that were gathered by making use of the Twitter data.
Once this was done, the researchers went on to the next step which involved matching the resulting simulations with the surveillance data generated by clinical and personal reports of influenza-like illnesses from the three countries and the US Centre for Disease Control (CDC). Their job was to deeply scrutinise the evolving dynamics subdued in the data. Using this, the researchers were successful in selecting a model that would be able to predict the future.
The next for the researchers involved taking the selected model and then testing it out against the official influenza surveillance systems. The results acquired depicted that the researchers' model accurately forecasted the disease’s evolution up to six weeks in advance, which is much earlier than what the other models available in the market could achieves.
The model is expected to enable public health agencies to plan much ahead in launching campaigns that encourage individuals to take preventative measures such as vaccination and increased hand washing and allocating needed medical resources.
According to Alessandro Vespignani, one of the researchers from the Northeastern University who invented the new model, in the past, humans had no information on the initial conditions for the flu. But, the initial conditions, which shows where and when did an epidemic start as well as the extent of infection, is actually very vital and functions as a launch pad for forecasting the spread of a disease.
Answering what sets their model apart from others already in the market, Vespignani notes that their explicit modelling of the infection's parameters, which involves details about the dynamics of the infection, helps in making the model unique from others. For example, with the model, researchers could identify the week when the flue epidemic would reach its peak and the magnitude of that peak with an accuracy of 70 to 90 per cent six weeks in advance of the event.
The key parameters of a disease helps in keeping a track of the seriousness of the flu in each year compared to the previous one and what was encouraging the spreading.
[Top Image: Reuters]
The new model developed by the Northeastern University researchers utilizes Twitter tweets to predict how a particular infection might affect the human population on planet Earth. The researchers set out gathering tweets along with the parameters of each of the season’s epidemic, such as how many people an individual with the virus can infect, the incubation period of the disease, viral strains present and the immunisation rate etc.
Once the tweets were gathered, they progressed to the next step which involved applying forecasting and a couple of other algorithms to the key parameters of the infection that were gathered by making use of the Twitter data.
Once this was done, the researchers went on to the next step which involved matching the resulting simulations with the surveillance data generated by clinical and personal reports of influenza-like illnesses from the three countries and the US Centre for Disease Control (CDC). Their job was to deeply scrutinise the evolving dynamics subdued in the data. Using this, the researchers were successful in selecting a model that would be able to predict the future.
The next for the researchers involved taking the selected model and then testing it out against the official influenza surveillance systems. The results acquired depicted that the researchers' model accurately forecasted the disease’s evolution up to six weeks in advance, which is much earlier than what the other models available in the market could achieves.
The model is expected to enable public health agencies to plan much ahead in launching campaigns that encourage individuals to take preventative measures such as vaccination and increased hand washing and allocating needed medical resources.
According to Alessandro Vespignani, one of the researchers from the Northeastern University who invented the new model, in the past, humans had no information on the initial conditions for the flu. But, the initial conditions, which shows where and when did an epidemic start as well as the extent of infection, is actually very vital and functions as a launch pad for forecasting the spread of a disease.
Answering what sets their model apart from others already in the market, Vespignani notes that their explicit modelling of the infection's parameters, which involves details about the dynamics of the infection, helps in making the model unique from others. For example, with the model, researchers could identify the week when the flue epidemic would reach its peak and the magnitude of that peak with an accuracy of 70 to 90 per cent six weeks in advance of the event.
The key parameters of a disease helps in keeping a track of the seriousness of the flu in each year compared to the previous one and what was encouraging the spreading.
[Top Image: Reuters]
Advertisements