Faster, cheaper, simpler: flow cytometry just got an upgrade
An AI-enabled microfluidic cytometer could make flow cytometry more accessible, affordable and faster.
A research team from George R Brown School of Engineering and Computing (Rice University; TX, USA), led by Peter Lillehoj and Kevin Mchugh, have developed an innovative AI-enabled microfluidic cytometer that can analyze cells in unpurified blood samples with similar accuracy to conventional flow cytometers. This small device could make flow cytometry more accessible, affordable and faster, providing intriguing possibilities for researchers working in remote locations, field researchers and for point-of-care clinical applications.
Flow cytometry, a widely used technique that sorts and analyzes cells in a fluid using a laser beam, is considered the “gold standard” for cell quantification and is extensively used in biomedical research. However, conventional flow cytometry techniques are time consuming, involve intricate sample preparation and require expensive and bulky equipment due to the specialized pumps and valves required for flow and control of the fluid. This limits the use of flow cytometry to high-resource laboratories with specially trained staff.
Attempts to make flow cytometry faster have been seen with impedance-based flow cytometers, which offer results in 5-20 minutes; however, the need for an expensive impedance analyzer limits its applications. Additionally, microfluidic cytometers that offer enhanced portability have been developed, but current platforms involve long sample preparation and processing protocols.
SLAS2025: innovation, automation and AI shape the future of life sciences
SLAS2025 once again kicked off the year as the hub for scientific collaboration and discovery.
To overcome these issues, the team developed an AI-based microfluidic cytometer with an innovative pump-free design that utilizes gravity-driven slug flow, allowing for a smaller more affordable device. Slug flow describes the movement pattern observed when a fluid composed of two fluids discrete phases passes through a channel and is a process usually observed in industrial applications. By recruiting gravity-driven slug flow, the team enabled samples to flow through the microfluidic device at a constant velocity, facilitating accurate cell identification and quantification and representing the “first time gravity-driven slug flow has been employed for a biomedical application,” according to senior author Peter Lillehoj .
In an illuminating test case for the technique, the team explored the platform’s ability to detect CD4+ T cells in fingerpick blood samples of healthy donors. Target CD4+ T cells in unpurified whole blood samples were labelled with anti-CD4 antibody-coated beads and driven through the microfluidic chip by gravity-driven slug-flow where they were recorded with a camera and optical microscope. The sample was then analyzed by a convolutional neural network-based model that had been trained to detect bead-labelled cells.
This experiment demonstrated that the cells could be detected with similar accuracy to conventional flow cytometry while being four times faster and less expensive. The investigation of immune cell subpopulations is an essential aspect of diagnostic applications and disease research that often relies on laborious flow cytometry techniques, making this a huge potential saving for researcher’s time and funding.
“Identifying and quantifying CD4+ T cells from unpurified blood samples is just one example of what one can achieve with this platform technology. This technology can be easily adapted to sort and analyze a variety of cell types from various biological samples by using beads labeled with different antibodies,” commented study lead McHugh.
By demonstrating the capabilities of this fast and affordable AI-enabled microfluidic cytometer, the team have highlighted the potential for its use in point-of-care clinical applications and low-resource settings. “Based on the promising results we’ve obtained so far, we are very optimistic about this platform’s potential to transform disease diagnosis, prognosis and the biomedical research landscape in the future” concluded McHugh.
