Of all the amazing technologies for seeing inside the body, one of the most versatile may be yet to come. Photoacoustic imaging (PAI) can produce instant three-dimensional images for diagnosis of diseases such as cancer and atherosclerosis. At Stanford, researchers led by electrical engineering Professor Butrus “Pierre” Khuri-Yakub are developing innovations to deliver PAI’s full potential as a multipurpose, sensitive, and safe imaging tool.

“In diagnostic imaging you want to detect the smallest, earliest, most minimal presence of a disease at a site in the body,” says Khuri-Yakub, who has focused on diagnostic imaging at Stanford for more than 30 years. “That’s really the motivation for wanting to optimize these technologies.”

In part because of advances Khuri-Yakub is pursuing in collaboration with students and colleagues at Stanford Engineering and at the medical school, PAI is beginning to look like it can offer many of the best attributes of many other diagnostic imaging technologies. Each existing technology has its limitations, so new technologies are still needed. For example, computerized tomography scans can penetrate deep into the body but involve exposing patients to a level of harmful radiation. Ultrasound has very high resolution but also produces a very artifact-riddled image.

But PAI has very high resolution (hundreds of millionths of a meter), penetrates as deep as three centimeters below the skin, produces a clear image, can distinguish among different tissues and is safe, Khuri-Yakub says. It can show, for example, where there is an abundance of oxygenated blood, which is often an indicator of tumors. Compared to some kinds of imaging, PAI is also relatively inexpensive.

Here’s how PAI works: Researchers shine fleeting pulses of infrared laser light on a sample, such as a cube of chicken meat. The light diffuses into the meat and heats tissues inside just a fraction of a degree. As they heat they expand, and when the laser pulse stops they cool and contract. Different tissues do this to different extents but their expansions and contractions produce shockwaves, which are also high-frequency sound waves. Those sounds are picked up by sensitive transducers in a detector. The detector feeds those signals to a computer which produces an image based on the patterns of emitted sound.

Microchips and nanoparticles
There are two opportunities to optimize PAI, and Khuri-Yakub’s group is pursuing both. One is to improve the transducers. To do that, Khuri-Yakub has cast aside the piezoelectric materials that have traditionally been used. Instead he makes arrays in silicon using techniques familiar in making microelectronics. Khuri-Yakub’s chips are called Capacitive Micromachined Ultrasonic Transducers, or CMUTs.

There are several advantages to rummaging in the semiconductor industry’s seemingly bottomless bag of manufacturing tricks. By making arrays of transducers in different geometric arrangements, for example, Khuri-Yakub can create sensors for a wide variety of tasks. An array of transducers produced on a flexible sheet of silicon could be wrapped around a catheter like the rubber grip on a pen. In that arrangement, the transducers can “see” on all sides of the catheter. If a doctor inserted a laser-equipped catheter into a major artery in a patient, the transducers could image the plaque buildup on the walls.

“We can easily make large-area 2D arrays, ring arrays, flexible arrays, and a lot of devices that are difficult if not impossible to make with traditional technology,” Khuri-Yakub says. In addition, the CMUTs can easily be integrated with silicon computer chips that can amplify and process the signals, says Ömer Oralkan, a postdoctoral researcher in the group. CMUTs can also be fashioned to be sensitive to a wider range of frequencies than can piezoelectric transducers. National Semiconductor has been making the chips for free for the group.

In a paper presented by doctoral student Srikant Vaithilingam at the October 2006 IEEE Ultrasonics Symposium in Vancouver, the group, including Vaithilingam and fellow student Ira Wygant, demonstrated several advances related to using CMUTs including their sensitivity to changes in tissue. The researchers took a piece of chicken meat and burned a small spot in the center. This intentional damage is called ablation, and the CMUTs imaged it very clearly. This is important because a treatment for abnormal heart rhythms (arrhythmia) is to precisely ablate the nerve tissue in the heart that carries the faulty signal. Imaging during the procedure is crucial because too little ablation won’t solve the problem but too much could damage the heart.

The other way to improve photoacoustic imaging is to augment the response of the tissue by injecting nanoparticle “probes” that have an enhanced response to the laser and produce stronger sound signals. The research was funded by the National Institutes of Health and Japanese imaging giant Canon Inc.

What makes nanoparticles—tiny gold balls 300 billionths of a meter in diameter—so special is that they can be fine-tuned in three important ways. First, researchers can attach proteins to them that are the exact match for proteins in tissues of interest, such as tumors. When such nanoparticles are injected, they will link up only with cancer cells. When they show up in a photoacoustic image, radiologists will be able to clearly see where the tumor is.

Second, researchers can enhance the nanoparticles to produce better image clarity. Materials science and engineering Assistant Professor Mark Brongersma, who is making the nanoparticles, can put a variety of chemical coatings on them to predispose them to absorb specific frequencies of laser light. They are then especially prone to heating and therefore producing sound.

Finally, the nanoparticles can also be crafted to resonate at particular ultrasound frequencies. This would make them especially well suited in systems that combine photoacoustic imaging with ultrasound imaging. Khuri-Yakub’s CMUTs can do such double-duty. Experiments using the nanoparticles as a photoacoustic “contrast agent” have begun in the lab of Sam Gambhir, a professor of radiology in the School of Medicine.

. While the research, both in nanotechnology and microelectronics, looks promising, the real benefit will come when the technology can reach patients. Says Vaithilingam, “I hope that our research at Stanford can be quickly translated into a clinical tool that doctors can use for diagnosing disease, monitoring treatment, and other urgent medical needs.”

Gambhir is optimistic: “Photoacoustic imaging when combined with molecular imaging strategies has the potential to dramatically change pre-clinical and clinical imaging, Methods for earlier diagnosis and better management of patients should result.”