Going Micro: Imaging Devices to Benefit Both Mouse and Biologist

Mice everywhere are breathing a collective sigh of relief. Soon, far fewer small laboratory animals will be routinely sacrificed to allow researchers to observe the after-effects of gene mutations or other experimental manipulations. Instead, investigators will be able to track the implications of those changes in living specimens in exquisite detail and for extended periods of time, thus painting a fuller, more accurate picture of what's going on and why. At least, that's the likely scenario if the various micro imaging and scanning technologies under development live up to their potential. Photo courtesy of Duke University's Center for In Vivo Microscopy Rat Livers: A series of MR microscopy images from the lab of Duke's Allan Johnson shows the effects of bromobenzene treatment on a rat liver over a period of 96 hours. Micro CAT (computed axial tomography) or micro CT, MR (magnetic resonance) microscopy, micro PET (positron emission tomography), and micro SPECT (single photon emission computer tomography) machines all have been in the works for the last several years. But high costs, limited accessibility, and less-than-adequate resolution have, for the most part, prevented commercialization and widespread use. Now, as more and more biologists realize the capacity of these machines in all-important areas such as cancer research, and as the equipment continues to improve, demand is starting to increase. The principle advantage of all micro imaging technologies is essentially the same: Scientists don't have to put knife to flesh in order to investigate the outcome of experimental changes made in small test animals. A series of high-resolution scans done at set intervals after a gene has been altered or cancerous cells injected effectively greatly increases the usefulness of each specimen. Photo courtesy of Brad Smith and Elwood Linney MRI Embryos: A series of MR microscopy images from the research of Duke's Brad Smith shows mouse embryo development from day 9.5 to neonate. The machines do, however, differ in cost, resolution, and specific applications. MR microscopy and micro CT devices, like their normal-sized clinical counterparts, are both intended primarily for scanning structural or anatomical features. Right now, MR offers better tissue contrast than micro CTs and, unlike its counterpart, no potentially lethal doses of radiation. "I would say that MR has a wider-spread application than micro CT," Allan Johnson, director of the Center for In Vivo Microscopy at Duke University, says of current systems. Like clinical MRI (magnetic resonance imaging) devices, MR microscopy uses radio- frequency energy to probe specimens, but with higher magnetic fields and specially sized receiver coils. A few companies already have sold a few hundred systems worldwide. Johnson, whose team has put together its own system adapted from a clinical General Electric device, emphasizes that the technology hasn't been used more widely because scientists expect the devices to be ready-made. "Researchers can buy a magnet, and the computer to run the magnet," explains Johnson. "But to make it run routinely, you also have to have other computers to do more sophisticated signal processing." Johnson, who heads a staff of 30 people, also notes the need for a coordinated effort among several specially trained personnel to carefully monitor specimens' vital signs. "It's something that developmental biologists are clamoring for--more ways to analyze the results of their genetic manipulations," explains Brad Smith, an assistant research professor in the department of radiology at Duke. Smith has used an MR microscopy device developed by Johnson to look at embryos and address developmental questions. He suggests that such technologies are likely to revolutionize his field. Photo courtesy of Oak Ridge National Laboratory Mouse Bones: 3-D skeletal image and surface rendering (inset) of an adult mouse with an obesity phenotype done with a MicroCAT developed by ORNL's Michael Paulus. New advances in the micro CT do, though, make it an attractive alternative to expensive MR machines. One recently unveiled micro CT prototype, due to be released commercially by ImTek Inc. of Knoxville, Tenn. later this year, boasts both high resolution and high throughput. With a projected price tag of $150,000, the device, called MicroCAT, is also a more feasible investment than an MR machine, which can cost anywhere from $500,000 to $1 million, depending on size and strength. Specially designed to do scans in mice in a matter of minutes, the MicroCAT, essentially a miniaturized medical X-ray CT scan for detecting soft-tissue abnormalities, can produce images showing details that measure as small as 50 microns. Developed by a team at the Oak Ridge National Laboratory (ORNL) in Oak Ridge, Tenn., it's the first micro CT machine able to scan live specimens. Unlike most miniature CT systems, the MicroCAT rotates around a stationary subject, rather than vice versa, to avoid a shifting of the animal's internal fluids. "To date, biologists have not, for some reason, been taking advantage of the tools that the medical practitioners have been taking advantage of for years: X-ray tomography and other structural imaging," says Michael J. Paulus, a researcher in the instrumentation and controls division at ORNL, and the team leader of the MicroCAT project. The ORNL team has used the machine to study fat deposits in a genetically engineered obesity-prone mouse. Declares Paulus, "It's very useful to be able to study the placement and the size of the fat bands without sacrificing the animal." Micro CT devices now available commercially, although only able to do scans on the body parts of dead mice, have already proven useful. Wes Beamer, a staff scientist at the Jackson Laboratory in Bar Harbor, Maine, and Mary Bouxsein, an instructor at the Beth Israel Deaconess Medical Center in Boston, have been collaborating to devise studies on mouse bone structure using a micro CT machine developed by Scanco, a company based in Zurich, Switzerland, that has plans of its own to release a device for scanning live specimens. Beamer and Bouxsein break specimens' bones to determine mechanical properties. Beamer envisions several applications for CT devices. How Well Do Mice Model Humans? See page 10 "NCI [National Cancer Institute] is going nuts because [the institute is] putting a lot of money into the different kinds of gene knockouts and transgenic studies," he says. "[But] then the ability to identify the phenotypes quickly and accurately turns out to be a huge problem." Bouxsein points out that the Scanco machines have also been of interest to pharmaceutical companies who want a way to examine the effects of drugs on animals without engaging in a cumbersome dissection process. Although Duke's Johnson thinks that MR, because of its images' superior tissue contrast, will ultimately be the most useful tool for scientists doing transgenic and molecular biology studies, he acknowledges that CT is "making inroads." His team has plans to build its own micro X-ray CT machine. Photo courtesy of UCLA School of Medicine PET Rat: A high resolution (1.8mm) whole-body image of a rat done with a MicroPET designed by UCLA's Simon Cherry illustrates high glucose utilization using a tracer. When it comes to imaging functional, metabolic processes such as glucose metabolism or brain receptor activity, however, the nuclear medical imaging devices, PET and SPECT, rather than MR microscopy or micro CT, are right for the job--despite their inferior resolution (currently 1.5 mm at best). Uses include imaging gene expression to discover when a particular gene is "on," how active that gene is, and the parts of the body where it's being expressed. Scientists can also use the devices to investigate how the brain responds to injury over time and how drugs and other forms of therapy affect the rate of recovery. "It's sort of become an interesting topic," says David Weber, the vice chair of radiology research at the University of California at Davis Medical Center, referring to his research into SPECT. "We introduced it about five years ago, but a curtailment in research funds set in and the interest was not high enough. [Now] interest is piquing." PET uses radioisotopes that emit positrons. The positron gets converted into two gamma rays that are emitted at 180 degrees apart and, based on the trajectory of those gamma rays, scientists can define the line along which the positron was emitted. With SPECT, only one gamma ray is emitted. The disadvantage: the device cannot indicate where the gamma ray comes from, so researchers have to place a collimator, essentially a piece of lead with some holes in it (hence the more accurate name, pinhole SPECT), in front of the detector to define the direction of the gamma rays. As a result, the number of detectable events is reduced, and the sensitivity of the device drops by roughly two orders of magnitude, compared to the micro PET's sensitivity. One big advantage for pinhole SPECTs: Single gamma ray radioisotopes are widely available because of their clinical applications and their longer half lives. Researchers don't need a particle accelerating cyclotron on site to make the necessary radioisotopes like they do with micro PET. Any laboratory doing routine clinical nuclear medicine--probably 5,000 laboratories in the United States alone--can simply buy a collimator, which costs from $500 to $3,000, attach it to their SPECT imaging system, and, after some alterations to their image reconstruction software, do high-resolution small-animal imaging. According to Simon Cherry, associate director of the Crump Institute for Biological Imaging at University of California at Los Angeles, despite the isotope issues, the outlook for micro PET machines is actually very good. He notes that at a recent workshop, of the 22 presentations, 20 were on micro PET and only two on pinhole SPECT. Cherry, whose team has designed a micro PET device (called MicroPET) to be commercially available from Concorde Microsystems by next year, points out that the increasing number of distribution centers set up for clinical PET machines have made the isotopes much more readily available. "But it's still an issue," admits Cherry. "A small lab that didn't have a clinical PET operation somewhere close by ... would probably want to go to an animal SPECT system." Still, Cherry does claim that Concorde has received a relatively high number of orders for the MicroPET, which costs $300,000 to $700,000, depending on the desired size. Cherry is also very interested in the capabilities of micro CT machines. He says, referring to Paulus, who's a friend of his, "I think both he and I have been very surprised by the response of biologists who are really keen to use these tools. So we're now working with [biologists] to improve [these tools] still further." In Vivo Microscopywwwcivm.mc.duke.edu/Brad Smith embryo.mc.duke.edu/animal/home.html Micro CT Oak Ridge National Laboratory www.ic.ornl.gov/rd-groups/msd/Projects/Microcat_Web/main.htm Micro PET UCLA www.nuc.ucla.edu/html_docs/crump/resprojects/microPET.html Pinhole SPECT UC-Davis www-radiology.ucdmc.ucdavis.edu/pinhole.html Society of Nuclear Medicine www.snm.org/ National Electrical Manufacturers Assoc. www.nema.org/ Of course, the use of one modality does not, by any means, preclude the use of another. Indeed, several teams are attempting to achieve the best of both worlds by marrying the micro PET's or pinhole SPECT's ability to detect functional, metabolic activities with micro CT's or MR's ability to detect structural or anatomical features. As part of a Small Animal Imaging Research Program, NCI has put out a request for applications; $13 million will be distributed among three or four U.S. facilities for the development of multimodality scanning systems for small-animal imaging. "There are a lot of people jockeying for that position," remarks Johnson, who was recently asked by the National Center for Research Resources (NCRR) to put together a small-animal imaging workshop. By the end of 1999, Paulus plans to combine X-ray CT and SPECT into a single prototype. Other teams are looking into PET-SPECT and CT-PET devices. "But I think that's a few years down the line yet," says Cherry. "We have to develop each separately and optimize that first." Right now, the best hope for combining the attributes of two modalities probably lies with the software rather than the hardware. By doing what's called image fusion or image registration, scientists can overlay and merge two or more images from different imaging modalities. Image registration capability is available now, but there's room for improvement. "It's something that's been slow in coming because of compatibility," maintains Weber. Incompatibility among instruments and the resulting inability to properly align coordinate systems from different machines has led the National Electrical Manufacturers Association (NEMA) to advocate DICOM (Digital Imaging and Communications in Medicine) standards for modality signals and output information. Nevertheless, Weber terms image registration "an exciting area." He expects significant advances over the next few years both in the clinical and the micro-imaging arenas.

Going Micro: Imaging Devices to Benefit Both Mouse and Biologist

Interested in reading more?

Become a Member of