Malaria breakthrough as AI could be used to predict next outbreak

Researchers at MBZUAI are using historical and real-time data to combat the disease

Efforts to stop the spread of malaria focus on mosquitoes, which carry the disease. AFP

A research team at Mohamed bin Zayed University of Artificial Intelligence (MBZUAI) is working on potentially groundbreaking research in the fight against malaria.

Using new technology and data techniques, the team can predict geographical areas vulnerable to the disease based on potential weather conditions, heat and humidity.

Led by Abdulmotaleb El Saddik, professor of computer vision at MBZUAI, the team is hoping to help doctors and health officials in Indonesia.

It is all made possible with sensory data fusion, a technique that combines data from several sensors to generate a virtual representation of the world, or a “digital twin”.

"The research we are dealing with is collecting sensory data from different sources and then performing machine learning algorithm and deep learning algorithm methods in order to predict potential weather conditions, potential areas depending on the heat, humidity and [in turn] predict potential malaria outbreaks,” Dr El Saddik said.

Malaria is a life-threatening disease primarily found in tropical countries.

About 250 million people contract it every year, with the mosquito-transmitted disease killing 600,000 in 2021 alone, according to the World Health Organisation (WHO).

What’s a digital twin?

“A digital twin is a virtual representation of any living or non-living entity, so, technically, a digital twin is our virtual representation of a city, forest or tree,” Dr El Saddik said.

“We perform interactions and we try to understand what's going on as if it was a real representation [of the world].”

Mohamed bin Zayed University of Artificial Intelligence and Korika, an Indonesian organisation, have teamed up to develop applications that could help predict malaria outbreaks. Photo: MBZUAI

Historical data from several sources is used to better represent the physical world and make sure that the information picked up by the AI sensors is correct.

Using a combination of historical and sensory data, researchers hope they can predict where the mosquitoes carrying malaria are most likely to migrate to next.

Why the need for sensors?

Dr El Saddik said it is necessary to use vision sensors and cameras to detect the location of mosquitoes.

“If we only consider satellite images, we can see the clouds where they are and where they are going. But for malaria, which is based on mosquitoes, distance from the ground is about two metres. So satellite images are not enough to give us this information,” Dr El Saddik added.

The team also uses data provided by meteorology centres that are used to teach the system AI algorithms.

Can it help against other diseases?

“We could use the whole framework to fight other diseases. However, we need to understand other diseases and we need the historical data from other diseases to train our artificial intelligence model to [imitate] them,” Dr El Saddik said.

“But once this project is done, we will have a much better understanding on the use of the data and we could apply our knowledge and train new models.”

Dr El Saddik said that in future, dengue fever could be another disease this technology is used to fight.