Doctors like to see what they are doing, and so they love magnetic resonance imaging scanners. MRI machines let them look inside the human body to see damaged tissues or dangerous tumors.
But MRI machines are designed in such a way that doctors cannot work on a patient while he is inside, to take tissue samples, for example, or even perform surgery. Now, researchers are developing tools to let doctors work on their patients while they are inside the MRI cocoon, a concept called “image-guided intervention.”
“It can offer very high precision and it can allow for less invasive surgery,” said Steven Krosnick, program director of image-guided intervention at the National Institute of Biomedical Imaging and Bioengineering in Bethesda, Md. “The surgeon knows immediately if they’re in the right spot, if the impact of their surgery is what it’s supposed to be.”
For example, physicians at Brigham and Women’s Hospital are testing a device that can be inserted inside an MRI machine and perform procedures via remote controls. For now they are using this so-called “robot” to take tissue samples of a man’s prostate as he lies inside an MRI scanner.
“We get live images, so you can watch the needle being inserted into the tissue,” said Gregory Fischer, associate professor of mechanical engineering and robotics engineering at Worcester Polytechnic Institute, who has spent six years developing the device.
Remotely controlled robots are already common in hospital operating rooms, where they are used in a wide variety of surgical procedures. But those are bulky machines that would never fit inside an MRI machine.
Standard surgical robots have iron and steel components that would wreak havoc inside the MRI; the machine’s intense magnetic field would yank such robots off their moorings and draw them deep into its interior.
“If there was a patient there, God forbid, there could be serious consequences,” said Krosnick.
Indeed, people have been injured or killed by flying metal objects left too close to an MRI. So Fischer and his team have developed a robot that is made mostly of plastic, powered by ceramic motors, and small enough to fit between the legs of a patient.
Prostate biopsies are usually performed “blind,” because the doctor cannot see the organ he is probing. As a result, doctors often take as many as a dozen tissue samples and still miss a tumor. An MRI-assisted biopsy eliminates this trial-and-error approach, by letting the doctor steer the sampling needle directly to a suspected tumor.
An MRI scanner is shaped like a large cylinder, with an opening at the center where the patient lies on his or her back. The machine bathes the human body in an intense magnetic field that shifts the alignment of the patient’s hydrogen atoms. Then, radio waves are beamed at the patient, causing some of the atoms to vibrate at a high frequency. This high-frequency vibration can be translated by computers into detailed pictures of tissues and organs.
MRIs are better than old-school X-rays at generating images of the body’s soft tissues. Besides, too much exposure to X-rays can be dangerous, whereas there is no evidence that the MRI’s magnetic field poses any health risks. But although the MRI makes it easier to see diseases, it also wraps around the patient’s body, so the doctor cannot reach the affected area while using the machine.
Fischer’s robot rests between the legs of a patient as he lies inside the machine. A doctor views a video screen that displays an MRI image of the prostate. With a remote control, the doctor will be able to precisely aim the biopsy needle.
If the biopsy tests prove successful, Fischer plans a version of the robot that could help treat prostate cancer by inserting radioactive “seeds” that kill cancerous cells. Such a system could avoid damage to healthy cells by putting the seeds only where they are needed.
Fischer earned his doctorate at Johns Hopkins University, and several colleagues from that school assisted in the development of his robot. However, a different team at the Johns Hopkins urology laboratory has developed a prostate robot of its own, and the US Food and Drug Administration has approved that one for clinical trials.
The Johns Hopkins robot is controlled using air pressure, an approach that Fischer rejected because the device is bulkier and more complicated to control. It also requires the patient to lie on his side rather than his back. Fischer said that this can result in distorted MRI images.
Instead, his team chose piezoelectric motors. These use ceramic compounds that vibrate under electric current. This vibration is translated into rotation like that of a standard motor, but the ceramic version contains no iron or steel, so it is unaffected by the MRI scanner.
Even so, the electrical signals to the motors threatened to scramble the MRI images. Fischer and his team used techniques similar to those found in high-end stereo systems to shield the MRI from any interference from the robot’s electronics.
Dan Stoianovici, the Johns Hopkins mechanical engineering professor who designed the air-driven robot, insisted his approach is best, because it can be used in any MRI machine, whereas the WPI robot might not work properly in machines with stronger magnetic fields. “They will be much more limited in terms of what they can be used for, and they will need much more testing,” he said.
But Fischer said that MRI machines do not vary that much. “It’s not like there’s hundreds of different scanners,” he said. “There’s maybe two or three,” and so compatibility testing will be a manageable problem.
MRI robots are not just for prostate cancer. Engineers at Vanderbilt University and at the University of Calgary in Canada have built remotely controlled devices for performing brain surgery.
Fischer is working on a similar device in cooperation with the University of Massachusetts Medical School. He has also collaborated with researchers at Harvard University’s School of Engineering and Applied Sciences to develop a robot that would be strapped to a patient’s abdomen and used to perform cancer surgery on lungs, livers, or kidneys.Hiawatha Bray can be reached at email@example.com. Follow him on Twitter @GlobeTechLab.