How AI is saving lives

16

September

2018

5/5 (1)

As the world becomes more connected, diseases are able to spread more rapidly. To prevent this from happening, the World Health Organization (WHO) has used various Epidemic Intelligence Tools over the last couple of decades. By monitoring various data sources, these tools try to detect future epidemics and monitor current outbreaks. These tools are becoming increasingly important to fight diseases and save lives. In 2017, the Epidemic Intelligence from Open Sources (EIOS) initiative was launched by WHO. The system was released on August 6th, 2018 and had already proven to be valuable.

Epidemic intelligence works in a continuous cycle of 8 steps. The first step is the collection of data from various sources. Examples of these sources are social media, national surveillance and laboratory networks. After the data is collected, it is processed to make sure that duplicate data is deleted and the data is linked to the correct diseases and locations. Thereafter, the system tries to detect patterns which suggest disease outbreaks. These possible outbreaks then need to be verified by lab confirmation, field verification or credible local sources. The situation gets analysed and the risk is determined. After making a visual representation of the current (and future) situation, action is taken if necessary. As new data is collected continuously, this cycle never stops.

The key feature from the new EIOS system is the large number of data sources. Over 8000 data sources are used to detect over half a million pieces information per month. This leads to 7000 signals, which are manually screened for relevance by the staff of WHO. After investigation, 100 events are investigated and 10 risk assessments are made each month.

During the first weeks that the EIOS system has been used, it has already detected several outbreaks which would have been detected in a later stage otherwise. For example, the cholera outbreak in Algeria and the leptospirosis outbreak in India have been detected in an early stage, enabling WHO to react quickly and save lives. These cases show demonstrate the potential to use data to improve global health.

Over the next period of time, WHO plans to further improve the EIOS system. One of the main priorities is to use speech-to-text technology to include radio sources to the system. This is especially valuable in poor regions, where radio still is one of the main methods of communication. Features like this could help WHO to further improve EIOS and save more lives.

 

 

This article is based on a presentation by Dr Oliver Morgan, Director Health Emergency and Risk Assessment – WHO Health Emergencies Program. Dr Oliver Morgan was one of the speakers at Intelligent Health 2018, an international congress on the use of artificial intelligent in health care. For more information about Intelligent Health, visit https://intelligenthealth.ai/

This presentation is not online accessible. However, a similar PowerPoint presentation can be found at: http://www.oie.int/eng/BIOTHREAT2017/Presentations/6.2_BARBOZA-presentation.pdf
Please note that this is not the version of that was used at Intelligent Health on September 13th 2018.

Article on cholera outbreak in Algeria: https://www.aljazeera.com/news/2018/08/algeria-person-dies-cholera-outbreak-180825173701802.html

Article on leptospirosis outbreak in India: https://www.news18.com/news/india/as-leptospirosis-threat-looms-after-floods-kerala-plots-30-day-micro-plan-1860783.html

Please rate this