Luke P. Lee, assistant professor of bioengineering at UC Berkeley, and his doctoral student Sunghoon Kwon have captured an image of a plant cell with a microlens smaller than the period at the end of this sentence.

"It's shrinking a million dollar machine down to a size that can balance on the tip of a ballpoint pen," said Lee, who presented the results at a recent International Conference on Micro Electro Mechanical Systems. "The microlens and scanner we've made is a crucial part of a microscope that is 500 to 1,000 times smaller than anything in its class."

In testing the accuracy of the microlens and scanner, Kwon placed a cell sample taken from a flowering lily, Convallaria majalis, onto the platform of a conventional confocal microscope. Without moving the sample, they captured a cross-sectional image of the cell wall, first with the traditional microscope, then with the microlens scanner. They found that the two images matched, showing for the first time that his microscopic lens could perform as well as a conventional one.

"Honestly, we were shocked," said Lee, who also is co-director of the Berkeley Sensor & Actuator Center. "What we've finally shown is a proof of concept. We have tested only 2-D images now, but it's just a matter of time and manpower before we get the first 3-D image."

The microlens and scanner are part of a device Lee is developing called the micro confocal imaging array, or micro-CIA. The micro-CIA belongs to a group of devices known as Bio-Polymer-Opto-Electro-Mechanical-Systems, or BioPOEMS. Invented by Lee, BioPOEMS marry the world of optics to that of microelectromechanical systems, or MEMS, for use in biological applications.

The size and sensitivity of the micro-CIA would allow technicians to quickly test even trace amounts of anthrax or smallpox in the field. It could become a crucial part of a "lab-on-a-chip," where researchers can study genes and proteins in ways unimagined decades ago. Lee is particularly excited by the potential for advancements in medicine possible with a miniaturized microscope.

"You could put this device on the tip of an endoscope that could be guided inside a cancer patient," said Lee. "Doctors could then see how tumor cells behave in vivo. It would also be feasible to deliver drugs directly to the tumor cell, and then view how the cell responds to the drugs."

High-end confocal microscopes, which house several lasers, take up to a meter of desk space, can cost more than $1 million and typically require highly-trained operators to run them, said Lee. The high cost of owning and running confocal microscopes limits the amount of research that can be done with them, he said.

"My goal is to not only shrink the size of these microscopes, but to make them as easy and as cheap to use as a digital camera," said Lee. It is with a hint of populist sentiment that Lee began devising a teeny version of the confocal microscope, the micro-CIA. He envisions a future where confocal microscopy is as common as a Bunsen burner in academic and industry research labs.

Unlike scanning electron microscopes, which construct 3-D topological images of dead cells, confocal microscopes can capture images of nanoscale activity inside living cells. Confocal microscopes also allow researchers to focus on specific components inside the cell, such as DNA strands, or mitochondria.

Cell parts marked with a fluorescent dye are "excited" by the laser and emit light back at specific wavelengths. Mitochondria, for instance, emit a fluorescent red color while nucleic acids emit a fluorescent blue, depending upon the molecular labeling of each component in the cell. To form 3-D images, 2-D slices are stacked together in a way similar to how an MRI image is formed.

Equipped with a microlens about 300 microns in diameter, the microscopic scanner Lee tested is a square of about 1 millimeter on each side and can move a distance of 50 to 100 microns. Lee is also testing a nanolens as small as 500 nanometers in diameter, or 200 times thinner than a strand of human hair, and smaller than the average red blood cell.

Lee's design of the micro-CIA will include three scanners stacked vertically above the staging platform where samples are studied. The scanners will measure each of the three axes - X, Y and Z - in three-dimensional space.

To make the scanner and lens, Lee employed technology similar to that used to manufacture microchips. The lens is made of a tiny drop of polymer shaped by surface tension and hardened by exposure to ultraviolet light. To focus the lens, Lee and Kwon adjusted the distance between the lens and sample. While it is also possible to focus by changing the shape of the lens, Lee said doing so would likely increase the cost and complexity of production, something he wants to avoid.

Comb-drives on each side of the microlens act as microactuators, tiny engines powered by electrostatic forces that move the microlens back and forth 4,500 times per second. Sensors then pick up fluorescent signals and feed the data back to a computer where the image is displayed in real time.

Lee's work is part of UC Berkeley's Health Sciences Initiative, which brings together scientists from disparate fields in the pursuit of major advances in health and medicine.

The research is part of a three-year project funded by the Defense Advanced Research Projects Agency.