Can AI identify mosquito species? VectorCAM, a pocket-sized device, uses machine learning to differentiate species with 95% accuracy, enhancing malaria surveillance efforts

Transcript

Not all mosquitoes are created equal. Of the more than three thousand species, only a limited number of the Anopheles genus can transmit malaria. Even within that subset, subtle physiological differences affect how malaria spreads. Some mosquitoes prefer to bite indoors, while others outdoors. Some need large bodies of water to breed, while others only need a small puddle. Distinguishing these species is critical for effective malaria control—whether using bed nets, indoor spraying, or outdoor larval management. But identifying them by eye takes expert, entomological knowledge. Could AI help? The VectorCAM team at Johns Hopkins is working on just that. Their pocket-sized device uses a small light and magnifying lens, allowing a phone camera to capture close-up images of mosquitoes placed on slides. With up to 95% accuracy, it can identify mosquito species based on morphology in seconds. The hope is that VectorCAM will help health teams better understand mosquito populations, paving the way for more targeted and relevant malaria control efforts.

Source

Towards transforming malaria vector surveillance using VectorBrain: a novel convolutional neural network for mosquito species, sex, and abdomen status identifications (Scientific Reports)

About The Podcast

The Johns Hopkins Malaria Minute podcast is produced by the Johns Hopkins Malaria Research Institute to highlight impactful malaria research and to share it with the global community.

For decades, insecticides have shielded us from malaria—but cracks are showing. Resistance is spreading, and environmental concerns are growing. Could a simple pouch of fruit juice with a powerful secret be the breakthrough we need?

with George Dimopoulos of the Johns Hopkins Malaria Research Institute

About The Podcast

The Johns Hopkins Malaria Minute is produced by the Johns Hopkins Malaria Research Institute to highlight impactful malaria research and to share it with the global community.

One of the main ways of controlling malaria is to reduce mosquito populations through insecticides. But the mosquitoes are developing resistance, making most insecticides less effective. What if the answer lies beneath our feet?

Transcript

One of the main ways of controlling malaria is to reduce mosquito populations through insecticides. But the mosquitoes are developing resistance, making most insecticides less effective. We need new vector control interventions – what if the answer lies beneath our feet? Researchers from the Dimopoulos Group at the Johns Hopkins Malaria Research Institute have turned to an unexpected source of inspiration—soil. They’ve produced a natural biopesticide, derived from a type of bacteria found in soil called Chromobacterium. When you deliver this biopesticide through a sugar bait – which lures the mosquitoes to feed on it – it kills the mosquitoes, regardless of their resistance to insecticides. Additionally, at non-lethal doses, Chromobacterium can enhance the effectiveness of other insecticides, acting as a synergist, as well as making mosquitoes incapable of finding a human to feed on. These findings were first demonstrated in the lab, but have now been confirmed in enclosed field trials in Burkina Faso. It's hoped that this naturally-occurring insecticide could support vector control efforts to curb disease transmission.

Source

Chromobacterium biopesticide overcomes insecticide resistance in malaria vector mosquitoes (Science)

About The Podcast

The Johns Hopkins Malaria Minute podcast is produced by the Johns Hopkins Malaria Research Institute to highlight impactful malaria research and to share it with the global community.