Researchers at Nanyang Technological University have developed a compact, AI-enabled biochip capable of detecting genetic disease markers within just 20 minutes. The innovation combines nanophotonics with artificial intelligence to deliver rapid, highly sensitive analysis of microRNAs. These are key biomarkers linked to a wide range of diseases.
MicroRNAs are short RNA molecules that regulate gene expression and are increasingly recognized as critical indicators for conditions such as cancer, cardiovascular disease, neurodegenerative disorders, and metabolic illnesses. However, detecting them has traditionally been technically challenging, requiring time-intensive laboratory processes such as polymerase chain reaction (PCR), which can take several hours and involve complex sample preparation.
Faster, simpler molecular diagnostics
The newly developed platform, described in Advanced Materials, offers a significantly faster alternative. Using only a small drop of blood, the biochip can simultaneously detect multiple microRNA targets without the need for amplification or labeling. According to the researchers, this reduces detection time from hours to approximately 20 minutes, while maintaining high sensitivity and accuracy.
At the core of the technology is a nanophotonic chip containing thousands of nanocavities. These are microscopic light-trapping structures that amplify fluorescent signals emitted when microRNAs bind to specific probes. These signals are captured in a single image using an integrated camera system and subsequently analyzed by AI algorithms.
The system employs a deep learning model known as Mask R-CNN to automatically identify, classify, and quantify microRNA signals. This eliminates the need for manual interpretation, reducing variability and human error while enabling high-throughput analysis.
Detecting disease
One of the key advantages of the platform is its ability to detect microRNAs at extremely low concentrations, down to just a few molecules per sample. In laboratory tests, the system achieved more than 99 percent accuracy in identifying target molecules across multiple test channels.
The researchers demonstrated the platform’s capabilities by detecting three microRNAs associated with non-small cell lung cancer, miR-191, miR-25, and miR-130a, in human cell extracts. Importantly, this was achieved without the need for amplification steps, highlighting the efficiency of the nanophotonic design.
“By combining signal enhancement at the nanoscale with AI-driven image analysis, we can detect minute amounts of RNA across thousands of sensing sites in just minutes,” said Bowen Fu, first author of the study and PhD researcher at NTU’s Institute for Digital Molecular Analytics and Science.
Non-invasive screening
Lead researcher Associate Professor Chen Yu-Cheng emphasizes the broader ambition behind the technology. “Our goal is to develop a system that can rapidly and accurately measure multiple biomarkers simultaneously,” he explained. “In the future, this could enable screening for hundreds or even thousands of disease markers using minimally invasive samples such as blood, saliva, or urine.”
Such capabilities could support large-scale screening programs and accelerate the shift toward personalized medicine, where treatment decisions are guided by individual molecular profiles. The research team has already developed a working prototype that integrates a color camera and a mobile application. The app processes captured images using AI algorithms and delivers results in near real time, suggesting potential for point-of-care or even decentralized diagnostic use.
Clinical and research implications
Experts see significant potential for clinical adoption, particularly in oncology. According to Associate Professor Sunny Wong Hei of Tan Tock Seng Hospital, technologies capable of accurately detecting multiple microRNAs could address a major unmet need in early disease detection and monitoring.
“MicroRNAs have long been recognized as promising non-invasive biomarkers, especially in cancer,” Wong noted. “A platform that enables rapid, precise detection could improve early diagnosis, patient stratification, and treatment monitoring.” Beyond clinical diagnostics, the technology may also find applications in pharmaceutical research, particularly in the development and testing of microRNA-targeted therapies.
Next steps
While the results are promising, further validation in clinical settings will be required before widespread implementation. The researchers are currently exploring the platform’s performance in more complex biological samples and investigating its scalability for broader diagnostic use.
Nevertheless, the study highlights the growing convergence of nanotechnology and artificial intelligence in healthcare. By enabling faster, more precise molecular analysis, innovations such as this AI-powered biochip could play a key role in advancing early detection and personalized treatment strategies across multiple disease areas.
Open-source software
Last year, researchers at the University of Navarra developed RNACOREX, an open-source software tool designed to uncover complex genetic networks in cancer. The platform identifies regulatory interactions between microRNAs (miRNAs) and messenger RNA (mRNA), supporting survival analysis and precision oncology. Developed at the DATAI institute and validated using data from The Cancer Genome Atlas, the tool addresses a key challenge in cancer research: translating large-scale genomic data into meaningful insights.
RNACOREX analyzes thousands of molecular interactions and ranks the most biologically relevant ones, enabling more accurate and interpretable predictions of patient outcomes. While matching the performance of advanced AI models, it stands out by offering transparency in how predictions are made. Freely available, RNACOREX supports clinicians and researchers in identifying actionable biomarkers and advancing personalized cancer care.
