Image created by Dr. Michael J. Miller
Combining microfluidics, Raman spectroscopy and deep learning, AutoEnricher is capable of rapidly flagging very low concentrations of pathogens and diagnosing multiple infections simultaneously. The new system has the potential to be a powerful tool in the fight against antimicrobial resistance (AMR), one of the world’s biggest emerging health challenges.
AMR is driven in large part by the overuse of antibiotics, particularly in emergency situations where the source of infection cannot be easily identified. Resistance to antibiotics is estimated to have caused five million deaths in 2019, a figure projected to rise to 10 million a year by 2050. By rapidly diagnosing the cause of infections, AutoEnricher could prevent the overuse of antibiotics that is fuelling the AMR crisis. The device is described in Nature Communications.
“One of the major drivers of antibiotic resistance is the misuse or overuse of drugs to treat infections,” said Jiabao Xu of Glasgow University's James Watt School of Engineering, one of the paper’s first authors.
“Currently, it can take days or even weeks to culture microbes taken from patient samples in the lab to enable diagnosis. That means doctors often have to act urgently and use antibiotics to treat patients suffering from life-threatening conditions like sepsis or pneumonia without knowing for sure if they actually have a bacterial infection.”
AutoEnricher works in two stages. In the first, a proprietary microfluidic device scrubs human cells from samples of blood, urine or spinal fluid, leaving behind only pathogen cells. The second stage uses Raman spectroscopy to determine the unique chemical fingerprints of the pathogens present. A deep learning model - trained on a database of 342 clinical isolates from 36 species of bacteria and fungi – then delivers a diagnosis in less than 20 minutes.
The performance of the system was validated with the help of three hospitals in China that provided samples from a total of 305 patients. These samples were also tested using conventional lab methods to culture the bacteria. AutoEnricher’s diagnosis matched the lab results 95 per cent of the time, while also managing to identify mixed infections that were missed in the lab.
“These are really encouraging results from the largest study of its kind conducted on real patient samples,” said Professor Wei Huang of Oxford University, a co-investigator on the project.
“We’ve shown that this single-cell approach to diagnosis can rapidly deliver remarkably accurate results, and even pick out multiple infections which are much harder to spot using conventional lab culture methods.”
Reference
Li, Y., Xu, J., Yi, X. et al. Rapid culture-free diagnosis of clinical pathogens via integrated microfluidic-Raman micro-spectroscopy. Nat Commun 17, 283 (2026). https://doi.org/10.1038/s41467-025-66996-y
Abstract
Antimicrobial resistance (AMR) is a critical global health challenge, demanding rapid and accurate diagnostics to guide timely antimicrobial therapy. Current diagnosis is hindered by prolonged culturing and difficulties detecting low pathogen loads. Here, we present a culture-free diagnostic platform that integrates microfluidics, Raman micro-spectroscopy, and deep learning to deliver “sample-to-report” testing within 20 min. The microfluidic enrichment system employs dialysis-dielectrophoresis (DEP) technology to rapidly isolate pathogens directly from clinical samples with a detection limit as low as <2 colony forming unit (CFU)/ml. Combining a single-cell Raman fingerprint database of 342 clinical isolates from 29 bacterial and 7 fungal species with a 1D ResNet deep learning model, our approach achieved 95.1% accuracy in lab settings. Validated in a 305-patient clinical study involving primary urine and other clinical samples, it demonstrated 95.4% agreement with traditional culture methods and 98.5% sensitivity in diagnosing infections. While broader validation is needed for clinical implementation, the integrated, rapid diagnosis pipeline, as well as broad-spectrum detection, offer a promising solution for next-generation diagnostics for combating AMR.