Researchers at Korea University have developed an ultra-sensitive HIV diagnostic platform that detects infection using a single drop of saliva [1].
This development is critical for early detection when antibody levels are low, a period where conventional rapid tests often fail to identify positive cases [1, 3]. By removing the need for blood draws, the platform could lower barriers to testing and prevent the spread of the virus.
The system utilizes nano-trap antibody capture to isolate the virus. This process allows for signal amplification of approximately 20-fold [1], increasing the visibility of the virus even in small samples.
To interpret the results, the researchers integrated a smartphone-based AI read-out. Park Jung-soo, a PhD student of the Mechanical Engineering Department, said the AI analyzes weak signals that are not easily visible to the naked eye to automatically determine positive or negative results [1].
The research team said the AI reading system demonstrated a diagnostic accuracy of 98.6% [1]. The platform was designed to capture HIV in its early stages using only saliva [2].
The announcement of the research was made on May 7, 2026 [2, 3]. The team's approach combines nanotechnology with machine learning to bridge the gap between laboratory-grade precision and point-of-care accessibility.
“The AI analyzes weak signals that are not easily visible to the naked eye”
The shift from blood-based to saliva-based testing represents a significant move toward non-invasive screening. By combining signal amplification with AI interpretation, this platform addresses the 'window period' of infection where viral loads are too low for standard rapid tests, potentially reducing the number of undetected carriers and improving public health outcomes.
