Gold mining in Venezuela’s Bolivar state is causing serious health problems. As forests are cut down for gold, more mosquitoes are spreading malaria. This is especially bad because Venezuela was declared malaria-free in 1961. Now, the disease is coming back fast, especially in remote villages where healthcare is hard to reach.
Malaria is still a big problem worldwide. In 2023, there were about 263 million cases and nearly 600,000 deaths, according to the World Health Organization (WHO). In Venezuela, many small communities don’t have the trained doctors or tools needed to check blood samples for malaria.
To help solve this, a group of researchers created a smart AI model using NVIDIA technology. The team — including Ramos-Briceño, Alessandro Flammia-D’Aleo, Gerardo Fernández-López, Fhabián Carrión-Nessi, and David Forero-Peña — trained a computer program to find malaria parasites in blood samples. The AI learned to detect two main types of parasites: Plasmodium falciparum and Plasmodium vivax.
They started with nearly 6,000 microscope images from a hospital in Bangladesh and used special software to cut and prepare the images. After editing and expanding the dataset, they had almost 190,000 clear images to train the AI. The result? The AI reached a very high accuracy of 99.51% in spotting the parasites.
To make this possible, the team used tools like PyTorch Lightning and NVIDIA CUDA, which helped process large amounts of data quickly. Compared to using a regular computer (CPU), this setup made the training much faster.
The study shows that AI could become a powerful new way to diagnose malaria, especially in places where expert help is hard to find.
Source: hitconsultant
