A $1.50 Lens-Free Microscope
The device could diagnose disease in the developing world and enable rapid drug screening.
Using a $1.50 digital camera sensor, scientists at Caltech have created the simplest and cheapest lens-free microscope yet. Such a device could have many applications, including helping diagnose disease in the developing world, and enabling rapid screening of new drugs.
The best current way to diagnose malaria is for a skilled technician to examine blood samples using a conventional optical microscope. But this is impractical in parts of the world where malaria is common. A simple lens-free imaging device connected to a smart phone or a PDA could automatically diagnose disease. A lensless microscope could also be used for rapid cancer or drug screening, with dozens or hundreds of microscopes working simultaneously.
The Caltech device is remarkably simple. A system of microscopic channels called microfluidics lead a sample across the light-sensing chip, which snaps images in rapid succession as the sample passes across. Unlike previous iterations, there are no other parts. Earlier versions featured pinhole apertures and an electrokinetic drive for moving cells in a fixed orientation with an electric field. In the new device, this complexity is eliminated thanks to a clever design and more sophisticated software algorithms. Samples flow through the channel because of a tiny difference in pressure from one end of the chip to the other. The device’s makers call it a subpixel resolving optofluidic microscope, or SROFM.
“The advantage here is that it’s simpler than their previous approaches,” says David Erickson, a microfluidics expert at Cornell University.
Cells tend to roll end over end as they pass through a microfluidic channel. The new device uses this behavior to its advantage by capturing images and producing a video. By imaging a cell from every angle, a clinician can determine its volume, which can be useful when looking for cancer cells, for example. Changhuei Yang, who leads the lab where the microscope was developed, says this means samples, such as blood, do not have to be prepared on slides beforehand.
The current resolution of the SROFM is 0.75 microns, which is comparable to a light microscope at 20 times magnification, says Guoan Zheng, lead author of a recent paper on the work, published in the journal Lab on a Chip.
The sensor has pixels that are 3.2 microns on each side. A “super resolution” algorithm assembles multiple images (50 for each high-resolution image) to create an enhanced resolution image–as if the screen had pixels 0.32 microns in size. However, super-resolution techniques can only distinguish features that are separated by at least one pixel, meaning the final resolution must be at least twice the pixel size. This is why a .32 micron pixel size yields only a resolution of .75 microns.
Zheng’s technique uses only a small portion of the chip, allowing him to capture cells at a relatively high frame rate of 300 frames per second. This yields a super-resolution “movie” of a cells at six frames per second.
Using a higher-resolution CMOS sensor should allow an even better ultimate resolution, says Seung Ah Lee, another collaborator on the project. Lee wants to get the resolution up to the equivalent of 40x magnification, so that the technique can be used for diagnosis of malaria via automated recognition of abnormal blood cells.
Aydogan Ozcan, a professor at UCLA who is developing a competing approach, says that Zheng’s work is “a valuable advance for optofluidic microscopy,” in that this system is simpler, offers higher resolution, and is easier to use than previous microscopes. However, Ozcan says that the technique has limitations.
The microfluidic channel must be quite small, says Ozcan, which means the approach can’t be applied to particles that might vary greatly in size, and the channel must be built to accommodate the largest particle that might flow through it. Ozcan’s own lensless microscope does not use microfluidic channels, and instead captures a “hologram” of the sample by interpreting the interference pattern of an LED lamp shining through it. This method has no such limitations.
“From my perspective, these are complementary approaches,” says Ozcan, whose ultimate aim is cheap, cell-phone based medical diagnostic tools for the developing world.