South Sudan medics trial AI app to identify snakes and improve bite treatment

A Médecins Sans Frontières nurse at Ameth-Bek hospital in South Sudan administers antivenom to Kamundi Makieu Pingiel, who was bitten by a soldier snake (Photograph: Christina Simons/MSF)

The race to treat snakebite patients in time to save them could be eased by the development of software powered by artificial intelligence.

The medical charity Médecins Sans Frontières (MSF) is trialling AI snake detection in South Sudan using a database of 380,000 pictures of snakes to identify venomous species.

According to the World Health Organization (WHO), as many as 5.4 million people are bitten by snakes each year, of whom up to 2.7 million become seriously ill and 138,000 die from complications. Identification is a difficult but crucial part of treatment to ensure that rare and expensive antivenoms are only used when necessary.

Dr Gabriel Alcoba, an MSF medical adviser on snakebites and neglected tropical diseases, said: “Early results are promising; the AI sometimes identifies snakes even better than experts.

“I remember a time when we used photo albums to identify snakes in MSF hospitals. Medical staff would flip through pictures to figure out which snake had bitten a patient,” he added.

The app is being piloted at two MSF hospitals in South Sudan, where the numbers of people taken to hospital with snakebites are high. Between January and the end of July 2024, more than 300 snakebite patients had been treated in MSF medical facilities across the country.

When someone is bitten, medics encourage pictures to be taken at the time or for their staff to return to the site, with caution, to photograph the snake.

The photos are fed into AI-powered software to help identify the type of snake and what type of treatment is needed, even before the patient gets to hospital. Alcoba said the program’s accuracy could be developed further with more funding, research and better quality photos.

“Often, patients receive the wrong treatment because the snake isn’t correctly identified, or valuable antivenom is wasted on bites from non-venomous snakes, which can also cause serious side-effects. Antivenom is rare and extremely expensive, costing a patient anywhere from a month to a year’s salary,” said Alcoba.

David Williams, a WHO snakebite expert, said bites can cause a person to stop breathing, as well as kidney failure, tissue damage and fatal haemorrhages.

Rural communities are worst affected, said Williams, and many of the 240,000 people left disabled by snakebites each year are pushed into poverty by the cost of treatment and loss of income.

He said climate breakdown was causing increasing concern about snakebites, with recent flooding leading to a rise in incidents in South Sudan, Bangladesh, Nigeria, Pakistan and Myanmar.

Many of those countries do not have enough treatments – only 2.5% of what is needed in Africa is available – and a lack of regulation has led to sales of fake antivenoms that have eroded trust within many communities.

Source: https://www.theguardian.com/global-development/2024/sep/25/south-sudan-medics-trial-ai-software-to-identify-snake-and-improve-bite-treatment