Clinical tests that identify and count normal and bacterial cells in blood and other samples can tell investigators the source of a bacterial infection or help monitor the immune health of people with HIV. A simple, lensless imaging system for doing so is in development by researchers at the University of California, Los Angeles. It uses a light-sensing processor, similar to ones in digital cameras, to count and quickly identify different cell types in blood and drinking water. Simple algorithms then identify and count the cells.

Expensive flow cytometers do most cell counting now but the technique requires multiple steps in a lab. "Screening for a few bacterial cells in a few milliliters of water takes hundreds of tests with a conventional microscope," says developer and UCLA assistant professor of electrical engineering Aydogan Ozcan. His imager, however, can spot a small numbers of cells in a large sample of water or blood.

The device records diffraction signatures, while conventional microscopes take detailed pictures of small samples. But the diffraction images contain enough information to identify and count cells.

Light passing through a cell diffracts or bends in a characteristic way. Each cell type produces a diffraction pattern that depends on its size, shape, and a refractive index. A library of characteristic diffraction signatures, compiled by Ozcan, identifies the cells. After capturing an image, the processor counts the number of cells in the sample. The calculations require little processing power.

Ozcan says the device can image as many as 100,000 cells in a 20 cm2 field of view in one second. In tests, the counter has determined the concentration of red blood cells in an unprocessed blood sample with 90% accuracy.

In the accompanying photo, a cell counter uses a digital camera’s image processor to record diffraction signatures from cells in blood and other samples. Simple algorithms identify and count the cells. Yeast cells are circled in blue, red blood cells are circled in red, and beads are circled in green.