Online Issues
<< All Back-issues
<< This Issue's Table of Contents
ILAR Journal V42(3) 2001
Impact of Noninvasive Technology on Animal Research
Challenges
Challenges in Small Animal Noninvasive Imaging
Robert S. Balaban and Victoria A. Hampshire
| Robert S. Balaban, Ph.D., is Scientific Director of the Laboratory Research Program, National Heart Lung and Blood Institute, National Institutes of Health, Bethesda, Maryland. Victoria A. Hampshire, V.M.D., is Director of Advanced Veterinary Applications, Bethesda, Maryland. |
Abstract
The current status and challenges of small animal noninvasive imaging are briefly reviewed. The advantages of noninvasive studies on living animals versus postmortem studies are evaluated. An argument is advanced that even in postmortem situations, noninvasive imaging may play an important role in efficiently characterizing small animal phenotypes as well as pathology. Issues of data interpretation under anesthetized conditions in live animal studies are also reviewed. Five imaging technologies are discussed briefly: magnetic resonance imaging and spectroscopy, ultrasound, computer-assisted tomography, positron emission tomography, and optical imaging. The structural and physiological information content of these different modalities is reviewed along with the ability of these techniques to scale down for use in small mammals such as mice and rats. In general, it was found that most of these technologies scale favorably to the study of small mammals, generally providing more physiological information than when used on the larger human scale. This finding suggests that these types of small mammal imaging capabilities will play a very significant role in the full utilization of these important animal models in biomedical research.
Key Words: anesthesia; CT; mouse; MRI; optical microscopy; PET; rat; ultrasound
Introduction
Small mammals, namely mice and rats, play an important role in biomedical research. These models are desirable due to low cost of maintenance and housing, short reproductive cycle, availability, and relative ease of transport. Over the last decade, a dramatic increase in mouse utilization has occurred due to the ability to modify the genotype of this animal rapidly. The mouse genotype can be manipulated almost at will, providing a unique tool in evaluating the effects of targeted manipulations on the phenotype of a mammalian system. This study of the "functional genomics" of the mouse will clearly be a major topic of biomedical research over the next decade. However, the methodology to evaluate the physiology or phenotype of the mouse and other small mammal models is still developmental at best. In many cases, a gene is removed, or knocked out, overexpressed, newly expressed, or mutated in an animal with a given hypothesis with regard to the eventual phenotype. Regrettably, it is rare that these genetic modifications have the desired effects on the mouse, and many unexpected phenotypical consequences are realized. This result is simply due to our ignorance in how genes are used in the development and function. Thus, this type of genetic manipulation is quickly being realized to be more an exploratory or discovery-based process rather than a pure hypothesis-driven experiment, which makes the development of screening techniques for the evaluation of animal phenotypes even more critical. By screening, we mean both the ability to evaluate numerous structures and organ systems simultaneously without necessarily targeting one system and the ability to look at large populations of animals frequently required in genetic studies, especially in the area of mutagenesis. Due to the inherent surveillance nature of most noninvasive imaging techniques, these approaches are ideal tools for discovery in evaluating the phenotype of a mouse as they are for discovery of disease in a human patient. In this article, we discuss the application of imaging as a small mammal phenotyping tool with a major focus on the mouse. We focus on the mouse because it will be the dominant mammalian model over the next 5 yr, and its diminutive size provides the greatest challenges and opportunities in the imaging sciences.
When considering using an imaging modality for small mammals, one must consider the changes in scale on both the overall size of the subjects as well as the physiologically relevant imaging volume. For example, in visualizing the resistance arterioles or capillary network of the skeletal muscle, the scale of the target vascular system is the same (150-10 mm). However, the skeletal muscle structures can be several orders of magnitude less when comparing mouse and man. This is a case in which the physiologically relevant imaging volume is nearly fixed between mouse and humans. Other examples of constant physiological volumes are individual neurons or ganglia. In contrast, an important measurement in the heart is the distribution of work and blood flow across the heart wall, or the so-called transmural distribution. In a human, this distribution requires a spatial resolution on the order of 2 ´ 2 ´ 7 mm, whereas in the mouse heart, the resolution must approach 0.2 ´ 0.2 ´ 1 mm. In this case, an increase in resolution volume of several orders of magnitude is required, similar to the overall scale of the animal. Thus, depending on the questions addressed, the experiment must take into account the required resolution for the physiological or anatomical measurement and not simply the scale of the animal alone.
The gross scale of the animal and its organs also influence the type and effectiveness of the imaging techniques. Many imaging techniques scale very well with size, permitting higher resolution and signal to noise ratio (SNR1) as the sample size becomes smaller. One major example of this influence is optics, which can be used to probe the entire adult mouse, but is relatively ineffective at viewing anything in humans except very superficial structures due to its poor penetration. For each of the modalities described, a general discussion on the effect of sample scale is presented.
Beyond the consideration of imaging modality, we have found that the basic question of whether living or postmortem studies should be conducted has a significant impact on the effectiveness of the study in terms of cost, time, quality, and, ultimately, scientific value. Based on this observation, a brief comparison of living versus postmortem studies is presented.
Postmortem Noninvasive Imaging
Naturally, when one considers a noninvasive technique, the advantages of performing studies on a living animal come to mind. However, one of the first issues to address when considering an imaging procedure is whether the study should be conducted on a live animal or postmortem. In clinical studies many times the definitive phenotype, or disease, is defined in an autopsy. Noninvasive imaging of morphology can be conducted at the highest level in a cadaver because no physiological motions are present and imaging time is not as critical. Thus, a well designed system for acquiring as much information as possible from postmortem animals may provide many investigators with the morphological and biochemical information they require in a timely and cost-effective manner. Noninvasive techniques are also advantageous compared with conventional gross pathology procedures, which require sectioning the animal and organs. This advantage is realized in cost, speed of acquiring data, and the nondestructive technology permitting follow-up postmortem studies as required. The nondestructive autopsy approach is even gaining favor in the evaluation of human and animal cadavers (Boyko et al. 1994; Ros et al. 1990). Technical development in this area could greatly improve the throughput and quality of these examinations and permit high-resolution (100-m m isotropic resolution) whole mouse studies on the order of 1 min, seriously competing with many gross pathology studies. The concept that a full three-dimensional image of the soft tissue structure, bones, and vasculature of an animal could be collected in a few minutes and delivered to the investigator's computer for analysis without destroying the animal for further study has many advantages.
A major challenge of this approach is the development of the high throughput systems to evaluate hundreds of animals for screening purposes. With postmortem subjects, one could imagine several robotic solutions to feed magnetic resonance imaging (MRI1) or computer-assisted tomography (CT1) devices with scores of animals in series or in parallel. Such devices are in development around the world. In addition, the automation of image interpretation and analysis must keep up with this data flow as well as just the ability to store and transfer these large data files. For example, a single whole mouse image at a 50-mm isotropic resolution would contain approximately 5 ´ 108 numbers. This is a remarkable amount of data for a small animal and likely only obtainable on a postmortem study due to the time required and physiological motion interferences at this high spatial resolution.
Live Animal Studies
The major advantage of noninvasive studies is the ability to conduct studies on living animals without significant consequence to the animal or its physiology. However, live animal imaging studies are very difficult to perform because they generally require an anesthetized animal and animal technical support to monitor the animal throughout the procedure and recovery. In addition, the physiological motion, support issues, and limited time available for the scanning generally compromise the quality of the imaging data compared with postmortem studies. For example, current technology applied to MRI of the mouse heart requires at least 1 hr of scan time, not including the anesthetic induction and recovery time, whereas a similar postmortem scan providing information on the entire cardiovascular anatomy of the animal could be conducted in minutes. The reason for this difference is that the heart and lungs are moving in the living animal. Thus, the time available to collect data is severely limited because gating to the two physiological processes is required to freeze the motion of the heart. This is especially true when one is attempting to detect the dynamics of the heart, which requires many high-resolution images. In a postmortem study, the imaging time is nearly 100% in a properly conducted experiment, without the need to correct for flow or gate to physiological parameters that also must be measured in a highly precise fashion in living animal studies. In addition, large improvements in magnetic field shimming are also realized in the postmortem condition. All of these factors contribute to a decrease of a factor of 10 in time. This difference multiplied by several hundred animals becomes highly significant. Similar time constraints exist for other imaging modalities. All issues concerning motion, physiological status, and temperature may play a role in these studies. Due to the high cost and low throughput of these vital imaging techniques, these approaches should be reserved for those studies requiring this type of examination. Some of the types of studies requiring vital measurements are included in Table 1.
Living animal studies are the most difficult to perform, yet they provide the greatest amount of physiological information. For this reason, these types of noninvasive studies are the focus of the remainder of this discussion. For the different imaging modalities discussed below, the image information content and scaling issues remain basically the same between living and postmortem studies.
Anesthesia Procedures
In vital noninvasive studies of small animals, the use of anesthesia is a major challenge. Active restraint of animals is possible for ultrasound and some other modalities. However, the physiological effects or reproducibility of the physical and mental stress imposed on the animal is unclear, especially in cardiovascular studies. Thus, most studies must be conducted under anesthesia. Regrettably, the impact of this anesthesia requirement is that the imaging data are frequently more profound on the interpretation than the imaging experiment itself. This is especially true with transgenic animals in which the phenotype might be expressed as an enhanced sensitivity to anesthesia; with minimal changes, the phenotype is normal or wild-type.
Because monitoring physiological function is the goal of many imaging studies, from cardiovascular behavior to neurological function, a regime must be picked to minimize the impact on the function of interest. Ultimately, the regime must meet the requirements of the imaging modality both from the perspective of time required for the studies as well as physical access to the animal. Because anesthesia is such an important aspect of small animal imaging, we discuss this major challenge in the field in some detail below.
The small size of the mouse or other small mammal is particularly challenging for maintaining a stable anesthetic plane due to problems with mechanical ventilation and online measures/adjustments of physiological function. Of paramount importance in achieving exact details about physiological changes in models is the need to control homeostatic mechanisms such as fluid and electrolyte balance, blood glucose, and acid/base balance. The most successful anesthetic protocols incorporate considerations for the maintenance of normal packed cell volume, total protein, and osmotic shifts, as well as the minimization of respiratory and metabolic disturbances in pH and normal substrate utilization through the provision of glucose and electrolytes. Naturally, this incorporation requires dynamic monitoring of blood gases and chemistry during an experimental procedure, which is very difficult to obtain from a mouse due both to the difficulty in obtaining vascular samples from these small animals and to the small volumes one can collect without affecting the animal's fluid balance. Thus, not only is microsurgery required for these procedures to monitor the animals, and in some cases deliver imaging contrast agents (MRI, ultrasound, CT, and positron emission tomography [PET1]) or tracers (PET), but microanalytical techniques are also required for analysis of the blood samples.
Adequate warmth is of particular importance in small animals because surface area to body weight ratios and metabolic rate are almost 10-fold higher than larger mammalian species, making the extrapolation of large animal procedures to the mouse difficult. For example, the large surface-to-volume ratios of mice require nearly twice the amount of fluid supplementation during anesthesia than most large laboratory animals. In addition, this fluid must be preheated to prevent its contribution to hypothermia during the procedure.
With regard to ventilation, several significant anatomical and physiological considerations must be understood about rodents before one can fully understand acid/base and respiratory balance during imaging procedures. First, because of the relatively large gastrocoele in rodents, thoracic compression is an issue in causing ventilation/perfusion mismatching. Rodents should be positioned at a slight incline, head above tail, to allow maximal costal movement. Additionally, the plane of anesthetic is important, and apnea must be avoided as much as is possible. Investigators can modify standard electrocardiographic equipment with nonmagnetic leads for the purposes of monitoring heart rate to assess depth of anesthetic. Second, large animals tend to adjust minute volume by increasing tidal volume and decreasing respiratory rate per minute. Rodents, however, tend to compensate in rate. The selection of an ideal anesthetic preserves or forces rate stability and preserves maximal costal expansion. Dalkara and colleagues (1995) have described a good example of success in mouse intubation. Several descriptions of small animal ventilator systems for imaging studies have been published (Hedlund et al. 2000; Minard et al. 1998).
The simplest and most consistent anesthetic regime used in MRI and PET procedures usually involves inhalation anesthetics and spontaneously breathing animal models. In these procedures, fast acting induction and recovery inhalation regimes afford greater safety as well as stable physiological effects. Animals are rapidly induced in anesthetic chambers that double as the animal holder for the imaging experiment. In these chambers the animal is continuously bathed in a gas anesthesia. This closed environment also permits the control of the temperature of the animal via the flowing gas as well as the oxygen and carbon dioxide levels for physiological perturbations. These simple nonventilated animal protocols have been very successful especially in studies that focus mostly on structure and not physiological function.
Magnetic Resonance Imaging (MRI) and Spectroscopy (MRS1)
MRI and MRS are based on the detection of the oscillating magnetic field induced from a special set of nuclides that posse a net spin in the presence of a strong magnetic field. A detailed discussion of MRI and MRS appears elsewhere in this volume (Chatham and Blackband 2001). MRI generally refers to the determination of the distribution of one molecule, such as water or fat, within a tissue at high spatial resolution. MRS generally refers to maintaining the spectral information in the magnetic signals from the nuclides, which permits the determination of the molecules or metabolites containing a given nuclide. The collection of this additional information in MRS along with the fact that metabolites are generally at low concentration results in the MRS experiment having a low SNR. These combined effects make any images collected with MRS very poor in spatial and temporal resolution. MRI and MRS must be conducted in a strong homogeneous magnetic field, which requires a specialized magnet as well as receiver coils to detect the nuclide signals. Because the absorption of these oscillating magnetic fields is relatively low in biological tissues, the penetration of these signals is excellent in most studies. The detection of the naturally occurring nuclide 1H found in water and fats is usually used for MRI studies providing an adequate SNR to create images with submillimeter resolution in vivo. The MRI signal from water protons is rich in information about the physiology and function of tissues because it is the solvent of the cell with very little occurring without some impact on the magnetic properties of this molecule (Balaban 1998). This information includes a diverse amount of information on blood flow and oxygenation as well as macromolecular composition and motion, tissue structure, temperature, contractile activity, nerve and muscle fiber orientation, and edema. A few other nuclides are present in adequate concentration to be used in natural abundance MRI, including 23Na (Christensen et al. 1996; Winter et al. 1998; Wolff et al. 1990; Xia et al. 1996) (total Na distribution, some information on intracellular and extracellular distribution), 31P (Hsieh and Balaban 1987) (metabolite distribution and metabolic rates), and 2H (Eng et al. 1990; Ewy et al. 1988) (structural and biochemical information). These studies usually have much lower spatial and temporal resolution than 1H due to their relatively low concentration and magnetic signal intensity per mole compared with 1H. However, unique information can be obtained even on this spatial scale. Usually any tracer nuclide added to the animal is in a concentration too low to permit a reasonable MRI experiment to be conducted. The exceptions to this are inhaled superpolarized gases (Cremillieux et al. 1999; Viallon et al. 1999) (lung volume imaging, perfusion imaging) and 2H- (Ewy et al. 1988; Robinson et al. 1998) and 19F-labeled fluorocarbon blood substitutes (Zimmermann et al. 2000).
MRS maintains the spectral properties of the nuclides permitting the determination of the chemical species with in tissues. The nuclear magnetic resonance (NMR1) spectral properties of a given metabolite are a function of the local magnetic interactions within the molecule providing, in most cases, a unique spectral fingerprint. Using this fingerprint, investigators can determine the concentration of a given metabolite, noninvasively. In some metabolites, the physiological milieu (e.g., temperature, pH, or free Mg++ concentration) can modify the molecular interactions and resulting NMR spectral properties, providing unique information on these more global parameters. The lower concentration of these metabolites or ions generally results in poorer SNR and lower spatial resolution than conventional MRI procedures on water protons. MRS can be used on some natural abundance nuclides to good effect, monitoring several important metabolic reactions including 1H (Gadian et al. 1986; Nielsen et al. 1999; Sebrie et al. 1998) (e.g., lactate, fat, glutamate, choline, inositol, glucose), 13C (Artemov et al. 1998; Hassel and Brathe 2000; Peled-Kamar et al. 1998) (e.g., fats, glycogen), and 31P (Holtzman et al. 1997; in't Zandt et al. 1999; Lukes et al. 1997; Robinson et al. 1998) (e.g., ATP, creatine phosphate, glucose-6-phosphate, free phosphate). More extensive studies in MRS have been conducted with a number of tracer molecules permitting the analysis of the intracellular milieu, metabolite kinetics, drug distribution and metabolism, ion fluxes, and metabolite dynamics. These nuclides used as tracers in MRS include 133Cs (Schornack et al. 1997), 2H (Eng et al. 1990), 15N (Kanamori and Ross 1997; Meynial-Denis et al. 1997), 17O (Arai et al. 1990; Fiat et al. 1992; Ronen et al. 1998; Sibson et al. 1998), 13C (Bhujwalla et al. 1994; Pascual et al. 1998; Peled-Kamar et al. 1998; Sibson et al. 1998), 87Rb (Cross et al. 1995; Deslauriers and Kupriyanov 1998; Syme et al. 1990), and 19F (Hees and Sotak 1993; Mason et al. 1993; McSheehy et al. 1997).
Contrast agents in 1H MRI are usually metal-based agents (including free Mn) that modify the magnetic relaxation properties of water. This permits the elimination or enhancement of the water depending on the agent and detection scheme. Specifically, agents to enhance the vascular bed or distribute in the interstial space have been very useful in angiography (Bogdanov et al. 1993; Lauffer et al. 1998) as well as perfusion (Ikehira et al. 1988; Lyng et al. 1998), tumor detection (Maurer et al. 2000; Saini et al. 1995), and neuronal fiber tracking (Pautler et al. 1998).
As can be seen from just the number of nuclides that can be detected, the information content of MRI is remarkable. An incomplete summary of this information is included in Figure 1. From these data, one can gather information ranging from structure to the chemical composition of some elements. Physiological information of blood flow, oxygenation, and volume is available along with the metabolism that is supported by these processes. Finally, information on the extracellular and intracellular milieu, including ion concentrations, pH, and temperature, can also be obtained.
MRI and MRS methods have been successfully applied to the mouse and rat due to the advantageous scaling factors that occur in magnetic resonance (Dubowitz et al. 2000; Fayad et al. 1998, Jacobs et al. 1999; Maxwell et al. 1998; Slawson et al. 1998; Wiesmann et al. 1998). The mouse embryo has been extensively characterized (Smith 2000). Indeed, many of the pioneering biological applications of MRI were first performed on small mammals due to the availability of appropriate-sized magnets. Small animals also permit the use of small magnetic resonance receiving coils, which increase the sensitivity to the magnetic fields generated by the nuclides. In other words, the closer a coil can be physically placed to a target organ, the better the SNR of the measurement. This is analogous to the improvement in reception of a radio signal in your car as you approach the transmitter in your destination city. Smaller subjects also mean that smaller magnets with higher magnetic fields can be used. The SNR of the MRI experiment roughly increases linearly with the magnetic field when the sample noise dominates. At the time of this writing, mouse studies can be conducted on 11.7-T or even higher field systems in comparison to the 1.5-T systems used in humans (T = Tesla or 10,000 gauss earths field is 0.5 to 1 gauss). This means that an approximate factor of 10 or more increase in SNR can be realized using these small high-field magnets. However, the imaging experiment is a three-dimensional problem, and the cube root of this factor of 10 must be taken to evaluate the net effect of this increase in SNR on image resolution. For example, if an imaging voxel is 2 ´ 2 ´ 20 mm at 1.5 T, the voxel can only be reduced isotropically to ~0.9 ´ 0.9 ´ 9 mm at 11.7. Thus, both the magnetic field and coil proximity issues must be used to optimize the MRI experiment on a small mammal. It is our opinion that the reduction in coil size is the greatest gain in MRI/MRS experiments, especially when superconducting coils (Hurlston 1999; Miller et al. 1999) may be used to eliminate the coil as a source of noise that may result in a stepper increase in SNR with magnetic field.
Finally, the small size of mice also permits the use of very high-powered imaging gradients, which makes it possible to obtain >100 gauss/cm at very high switching rates. This permits the high spatial resolution required as well as a reasonable desired speed of acquisition without causing neuronal stimulation in these small subjects. This high rate of acquisition can contribute to increases in SNR via true fast imaging with steady state procession ("FISP"), with short time to repeat values and reduced in-homogeneity effects via shorter time to echo values. The sum of these advantages results in the acquisition of images in mice that are very comparable to those of humans, taking into account the physiologically important voxel size. Examples of a human and mouse heart are shown in Figure 2.
It is apparent from the progress in small animal MRI and MRS studies that much of the utility of these approaches in man will translate to the evaluation of small animal physiology.
Ultrasound
Ultrasound relies on the modification of an induced acoustic wave traveling through tissue. Ultrasound studies are conducted using a probe to project sound into the animal and recording the time and magnitude of the reflected sound wave using the same probe. The analysis of this acoustic echo permits the imaging and measurement of tissue acoustic properties. A detailed discussion of ultrasound appears elsewhere in this volume (Coatney 2001).
Ultrasound is used primarily for monitoring tissue structure and motion. The amount of tissue characterization that can be accomplished with ultrasound is limited. Because acoustic waves do not penetrate bone, ultrasound is not very useful in the developed brain with an intact skull. Another limitation of ultrasound is the fact that the acquisition of data is very user dependent in finding the appropriate acoustic "windows" for access of internal organs. Thus, a skilled technologist or investigator is required for these studies. Despite these limitations, ultrasound has been the mainstay in the evaluation of cardiac wall function, blood flow, and valve performance as well as an important tool in monitoring fetus development. The advantages of ultrasound include its ease of use, portability, and relatively low cost compared with MRI, CT, and PET. Due to its portability and the rapid frame rate that is relatively insensitive to motion, it is conceivable that nonanesthetized imaging studies could be conducted on restrained animals if the stress on the animal were not too great or influenced the results significantly. The portability and relative low cost suggest that these units could be housed in the animal holding facility permitting on-site data collection.
Recently, the development of bubble-based contrast agents (Calliada et al. 1998; Lindner et al. 2000) has improved the amount and quality of information in ultrasound imaging. Agents are now being developed for specific adhesion to useful markers in tissue that could expand the use of ultrasound in tissue analysis (e.g., Villanueva 1998). Another important contribution in this area has been harmonic-based ultrasound echo analysis (Forsberg et al. 1996, 2000), permitting higher tissue and agent-generated contrast. Real-time three-dimensional ultrasound (Linney and Deng 1999) performed on small animals may also remove some of the operator limitations as well as provide rapid whole animal studies. These approaches provide a significant opportunity in ultrasound in small animals to develop approaches to improve the specificity and physiology information content of this approach.
Ultrasound does scale appropriately with small animals. The resolution of the ultrasound experiment is roughly proportional to the frequency of operation. Smaller animals with a shorter path-length requirement for the ultrasound wave permit the use of higher frequencies that cannot adequately penetrate into larger animals or humans. The SNR of ultrasound is roughly constant, because the noise is a coherent "speckle" from internal reflections. However, as in magnetic resonance, the SNR of ultrasound is increased by the proximity of the probe to the region of interest. Thus, in smaller animals the target organ is much closer to the probe, which increases the inherent SNR. These factors permit the spatial resolution of the ultrasound instrument track with the size of the animal.
Again, using the heart as a comparison, the mouse images provided at 25 mHz and even higher frequencies are approaching the utility of human scale instruments. Numerous investigators have begun to use them in the evaluation of adult (Takeishi et al. 2000; Yokosawa et al. 2000) and fetal heart anatomy and function (Linask and Huhta 2000; Turnbull 2000). An example of an ultrasound image at 45 mHz is shown in Figure 3.
Compared with MRI, challenges in ultrasound imaging are tissue contrast and SNR. It is hoped that the advent of specific contrast agents will improve the utility of ultrasound beyond the structure/function studies now under way in animals. Regardless, ultrasound remains one of the least expensive and easiest to use of the imaging tools available for small animal evaluations concerning heart function and soft tissue structure outside the brain. Its ability to monitor fetal development under nonanesthetized conditions is also a valuable asset to the evaluation of development in transgenic animals.
CT
CT is basically a three-dimensional x-ray technique that is sensitive to the x-ray absorption of the tissue. Contrast can be generated by the differences in tissue absorption, with bone providing the most striking intrinsic contrast, or by using contrast agents to enhance the vasculature or specific tissues and conditions. The inherent SNR of CT is very high. Smaller animals provide some advantage in CT by permitting the use of low energy irradiation that can penetrate the mouse without significant attenuation. The low-energy irradiation is more sensitive to the tissue absorbance, which provides a higher contrast image. The use of low-energy irradiation in larger animals is prohibited due to the power deposition required to obtain an adequate flux through large structures. Finally, the size of the device can be greatly reduced, decreasing price, ease of shielding, and siting within an animal facility. Using current technology, full three-dimensional mouse images with 100 ´ 100 ´ 100 mm resolution can be obtained in a few minutes (Graichen et al. 1998; Kennel et al. 2000; Paulus et al. 2000; Yamashita et al. 2000) with higher resolution studies approaching potentially 50-m m isotropic resolution with the one limitation being that the amount of energy absorbed by the animal may approach "invasive" levels. A coronal CT section of a mouse is shown in Figure 4. The skeletal system highlighted in this CT image has been the focus of most of the initial work with small animal CT.
The high speed and high resolution of CT will clearly make it a valuable tool in the screening of large mouse populations. High-throughput systems for both vital and postmortem studies must be interfaced to these scanners to expedite the population screens envisioned for functional genomics. Contrast agents to improve soft tissue contrast as well as directly observe the vascular anatomy may also prove useful in the optimization of this approach to small animals, again requiring vascular access in the small animal. The major limitation of CT for use in small animals is the lack of information on tissue characterization and physiological function. Of all of the methods, CT will likely provide the greatest challenge in terms of data processing and data interpretation. As discussed above, a single data set from an isotropic 50-m m CT scan will approach 10 9 values. Due to the rather inefficient image reconstruction algorithms available for CT, these data will require a considerable amount of time just to convert into a useable image, not including any image processing or artificial intelligence to automate the interpretation of the images.
PET
PET relies on detection of radioactive probes emitted in the body. Imaging of this emission is performed using a combination of detector geometry along with the timing of the emissions detection. A detailed discussion of PET appears elsewhere in this volume (Cherry and Gambhir 2001). PET is one of the most sensitive imaging techniques and is capable of detecting vanishing small amounts of radiolabeled material. The short-lived isotopes (Ingvar et al. 1991) used in this approach include 11C (Kuge et al. 2000; Levchenko et al. 2000), 13N, 15O (Magata et al. 1995; Yamamoto et al. 2000), and 18F (Ingvar et al. 1991; Yamamoto et al. 2000) isotopes, which are extremely useful in the evaluation of biological processes. The clever use of these PET tracers and tracer chemistry is rivaled only by MRI/MRS in information content in an imaging modality (Phelps 2000). In addition to flow and metabolism markers similar in both PET and MRI, the sensitivity of PET has resulted in a unique ability to monitor receptor ligand interactions in humans and animals with remarkable success. This sensitivity has resulted in PET being one of the primary targets in the development of gene expression markers as well as the detection of early cancer (Phelps 2000). One of the major drawbacks of PET is the requirement for a local cyclotron to generate the probes and synthesis unit to produce the biologically useful probes. However, most major medical centers already have such facilities where the tiny quantities required for small animal imaging can be easily obtained. Because radioisotopes must be used in these studies, vascular access or direct injection of the tracers into the organ of interest is required.
PET has been demonstrated in the mouse and other small animals (Chatziioannou et al. 1999; Gambhir et al. 2000; Yu et al. 2000) using devices that can generate >5-fold higher overall spatial resolution than conventional human scanners (Cherry et al. 1997). For receptor binding in the brain, where the density of receptors may be similar in mice and humans, this resolution may still be inadequate for some studies. The major improvements in these small PET systems include the ability to design detectors that could more completely enclose the entire animal and collect a higher fraction of the emitted photons. In addition, for a given radioisotope, the small size of the animal results in less scattering and attenuation of the photons resulting in higher collection efficiency. These combined effects have resulted in very successful application of PET to the study of small animals. An example of the detection of reporter gene (herpes simplex virus 1 thymidine kinase) in the mouse using a 18-F labeled metabolite is shown in Figure 5 (Gambhir et al. 1999) from the UCLA group.
With these early demonstrations in small animals and the commercial availability of high-resolution PET instruments, the use of PET in the characterization of mice will certainly increase. The major advantage of PET is high sensitivity without the penetration limitations of optical techniques. The successful application of this approach will depend on the development of appropriate probes that will determine the specificity and sensitivity of the measurements.
Optical Imaging
Optical imaging is an extremely sensitive measurement that can detect a single molecule using fluorescence techniques. Optical imaging is usually performed in two modes: simple transmission absorption imaging and fluorescence imaging. In simple transmission absorption imaging, either transmitted or reflected light is used with tissue or optical probes, providing differential absorption to generate useful tissue contrast. The most common technique used here is optical coherence tomography (Chen et al. 1998; Kehlet et al. 1999; Roper et al. 1998). Fluorescent imaging is performed by irradiating the tissue with a frequency of light lower than the emission frequency exciting fluorescence from intrinsic or extrinsic probes under investigation. Fluorescence imaging is the most sensitive approach, and it has gained great interest with the development of genetically encoded highly efficient fluorescent probes based on green fluorescence protein. Optical imaging in large animals has mostly been limited to the study of skin, eyes, surface vessels, and epithelial tissues accessible to visible light (Barton et al. 1999; Chen et al. 1998; Kehlet et al. 1999; Knuttel and Boehlau-Godau 2000; Masters et al. 1998; Roper et al. 1998). Some spectroscopic studies with limited imaging resolution have been conducted with infrared light (IR1). These studies have been mostly focused on imaging structures using the natural differential optical absorbance of tissues. Most notable have been the direct observations of blood cell motion using the high extinction of the hemoglobin. Some naturally occurring probes including myoglobin, hemoglobin, and cytochromes can provide biochemical information within the tissue specifically dealing with oxygen delivery and mitochondrial energetics (Steen et al. 1989; Steinberg et al. 1997; Ueda et al. 1988). However, most of the activity in optical imaging focuses on the use of exogenous or genetically engineered optical probes. Probes are being created to evaluate everything from structure to intracellular milieu to protein function in vivo with unparalleled sensitivity and spatial resolution. Like all newly introduced probes, the adverse effects of these molecular manipulations must be carefully followed to ensure that the measurement or probe is not interfering with the background phenotype or physiology (Huang et al. 2000).
The major limitation of light is the high absorption and scattering that occur in biological tissues and limit the penetration of the light through the body. However, in small animals the required path-length of light is much shorter, which makes the use optics much more feasible. Multiphoton (so-called two- or three-photon confocal microscopy) fluorescence microscopy has also improved the depth of resolution and quality of fluorescence imaging in intact tissues by limiting the excitation field. Using this latter approach with specific protein fluorescence probes, the morphology and plasticity of neurons have been directly observed in vivo using two-photon confocal microscopy in rats (Figure 6) (Lendvai et al. 2000). The feasibility of using this approach on entire embryos without significant physiological effects on development has also been demonstrated (Squirrell et al. 1999). On a larger scale, whole adult animal screening has recently been shown using whole body IR fluorescence (Weissleder et al. 2000). An example from this study is shown in Figure 7 in which a tumor that trapped a polymer containing a near-IR fluorescence probe was localized in the intact mouse.
The optical detection of molecular events is the most sensitive molecular imaging tool available in vivo. Due to the small size of the mouse and the ability to create optical markers for monitoring a wide variety of gene functions or even simple physiological and anatomical questions, it is clear that this approach will play a growing role in the noninvasive evaluation of the mouse phenotype.
Summary
Most of the standard imaging modalities used in clinical evaluations scale favorably to the size of a mouse or rat. These improvements in performance result in the maintenance of the physiologically relevant information in these images even though the size of the subjects has been reduced by several orders of magnitude. Due to its high sensitivity and specificity coupled with the ability to create genetically coded probes, optical imaging is likely to play a growing role in small animal imaging. Major challenges in this approach are the maintenance and monitoring of an appropriate physiological state while conducting these studies. All of the modalities must be further modified to optimize their performance in the study of small animals; however, as noted in this review good progress is being made. Potentially, the largest technical challenge involves handling and processing the enormous amount of data provided by the approaches. One of the advantages of studying small mammals is that large numbers can be evaluated for genetic or mutant screening. Imaging these large numbers will result in the mandatory development of computational systems capable of handling these large data sets. In addition, analysis of these data must be automated in some form to reduce the need for the investigator to screen each of these images. These automated image interpretation systems are also under development in the clinical radiology community where useful approaches may already be undergoing evaluation. With the growing interest in the function of genes in development and function as well as the study of intact biological systems in mammals, it is clear that these screening imaging tools will play a critical role.
1Abbreviations used in this article: CT, computer-assisted tomography; IR, infrared light; MRI, magnetic resonance imaging; MRS, magnetic resonance spectroscopy; NMR, nuclear magnetic resonance; PET, positron emission tomography; SNR, signal-to-noise ratio.
References
Arai T, Nakao S, Mori K, Ishimori K, Morishima I, Miyazawa T, Fritz-Zieroth B. 1990. Cerebral oxygen utilization analyzed by the use of oxygen-17 and its nuclear magnetic resonance. Biochem Biophys Res Commun 169:153-158.
Artemov D, Bhujwalla ZM, Pilatus U, Glickson JD. 1998. Two-compartment model for determination of glycolytic rates of solid tumors by in vivo 13C NMR spectroscopy. NMR Biomed 11:395-404.
Balaban RS. 1998. Physiological and biochemical information from water in cardiac MRI. In: Higgins CB, Ingwall JS, Pohost GM, eds. Current and Future Applications of Magnetic Resonance in Cardiovascular Disease. Armonk NY: Futura Publishing Inc. p 321-336.
Barton JK, Hammer DX, Pfefer TJ, Lund DJ, Stuck BE, Welch AJ. 1999. Simultaneous irradiation and imaging of blood vessels during pulsed laser delivery. Lasers Surg Med 24:236-243.
Bhujwalla ZM, Shungu DC, Chatham JC, Wehrle JP, Glickson JD. 1994. Glucose metabolism in RIF-1 tumors after reduction in blood flow: an in vivo 13C and 31P NMR study. Magn Reson Med 32:303-309.
Bogdanov AA, Jr., Weissleder R, Frank HW, Bogdanova AV, Nossif N, Schaffer BK, Tsai E, Papisov MI, Brady TJ. 1993. A new macromolecule as a contrast agent for MR angiography: preparation, properties, and animal studies. Radiology 187:701-706.
Boyko OB, Alston SR, Fuller GN, Hulette CM, Johnson GA, Burger PC. 1994. Utility of postmortem magnetic resonance imaging in clinical neuropathology. Arch Pathol Lab Med 118:219-225.
Calliada F, Campani R, Bottinelli O, Bozzini A, Sommaruga MG. 1998. Ultrasound contrast agents: basic principles. Eur J Radiol 27(Suppl 2):S157-S160.
Chatham JC, Blackband SJ. 2001. NMR spectroscopy and imaging in animal research. ILAR J 42:189-208.
Chatziioannou AF, Cherry SR, Shao Y, Silverman RW, Meadors K, Farquhar TH, Pedarsani M, Phelps ME. 1999. Performance evaluation of microPET: A high-resolution lutetium oxyorthosilicate PET scanner for animal imaging. J Nucl Med 40:1164-1175.
Chen Z, Milner TE, Wang X, Srinivas S, Nelson JS. 1998. Optical Doppler tomography: imaging in vivo blood flow dynamics following pharmacological intervention and photodynamic therapy. Photochem Photobiol 67:56-60.
Cherry S, Gambhir SS. 2001. Use of positron emission tomography in animal research. ILAR J 42:000-000.
Cherry SR, Shao Y, Silverman RW, Meadors K, Siegel S, Chatziioannou A, Young JW, Jones WF, Moyers JC, Newport D, Boutefnouchet A, Farquhar TH, Andreaco M, Paulus MJ, Binkley DM, Nutt R, Phelps ME. 1997. MicroPET: A high resolution PET scanner for imaging small animals. IEEE Trans Nucl Sci 44:1161-1166.
Christensen JD, Barrere BJ, Boada FE, Vevea JM, Thulborn KR. 1996. Quantitative tissue sodium concentration mapping of normal rat brain. Magn Reson Med 36:83-89.
Coatney RW. 2001. Ultrasound imaging: Principles and applications in rodent research. ILAR J 42:232-247.
Cremillieux Y, Berthezene Y, Humblot H, Viallon M, Canet E, Bourgeois M, Albert T, Heil W, Briguet A. 1999. A combined 1H perfusion/3He ventilation NMR study in rat lungs. Magn Reson Med 41:645-648.
Cross HR, Radda GK, Clarke K. 1995. The role of Na+/K+ ATPase activity during low flow ischemia in preventing myocardial injury: A 31P, 23Na and 87Rb NMR spectroscopic study. Magn Reson Med 34:673-685.
Dalkara T, Irikura K, Huang Z, Panahian N, Moskowitz MA. 1995. Cerebrovascular responses under controlled and monitored physiological conditions in the anesthetized mouse. J Cerebr Blood Flow Metab 15:631-638.
Deslauriers R, Kupriyanov VV. 1998. Cardiac magnetic resonance spectroscopy. Biochem Cell Biol 76:510-521.
Dubowitz DJ, Tyszka JM, Sewry CA, Moats RA, Scadeng M, Dubowitz V. 2000. High resolution magnetic resonance imaging of the brain in the dy/dy mouse with merosin-deficient congenital muscular dystrophy. Neuromus Disord 10:292-298.
Eng J, Berkowitz BA, Balaban RS. 1990. Renal distribution and metabolism of [2H9]choline. A 2H NMR and MRI study. NMR Biomed 3:173-177.
Ewy CS, Ackerman JJ, Balaban RS. 1988. Deuterium NMR cerebral imaging in situ. Magn Reson Med 8:35-44.
Fatkin D, Christe ME, Aristizabal O, McConnell BK, Srinivasan S, Schoen FJ, Seidman CE, Turnbull DH, Seidman JG. 1999. Neonatal cardiomyopathy in mice homozygous for the Arg403Gln mutation in the alpha cardiac myosin heavy chain gene. J Clin Invest 103:147-153.
Fayad ZA, Fallon JT, Shinnar M, Wehrli S, Dansky HM, Poon M, Badimon JJ, Charlton SA, Fisher EA, Breslow JL, Fuster V. 1998. Noninvasive in vivo high-resolution magnetic resonance imaging of atherosclerotic lesions in genetically engineered mice. Circulation 98:1541-1547.
Fiat D, Ligeti L, Lyon RC, Ruttner Z, Pekar J, Moonen CT, McLaughlin AC. 1992. In vivo 17O NMR study of rat brain during 17O2 inhalation. Magn Reson Med 24:370-374.
Forsberg F, Goldberg BB, Liu JB, Merton DA, Rawool NM. 1996. On the feasibility of real-time, in vivo harmonic imaging with proteinaceous microspheres. J Ultrasound Med 15:853-860.
Forsberg F, Shi WT, Goldberg BB. 2000. Subharmonic imaging of contrast agents. Ultrasonics 38:93-98.
Gadian DG, Proctor E, Williams SR, Cady EB, Gardiner RM. 1986. Neurometabolic effects of an inborn error of amino acid metabolism demonstrated in vivo by 1H NMR. Magn Reson Med 3:150-156.
Gambhir SS, Barrio JR, Phelps ME, Iyer M, Namavari M, Satyamurthy N, Wu L, Green LA, Bauer E, MacLaren DC, Nguyen K, Berk AJ, Cherry SR, Herschman HR.. 1999. Imaging adenoviral-directed reporter gene expression in living animals with positron emission tomography. Proc Natl Acad Sci U S A 96:2333-2338.
Gambhir SS, Herschman HR, Cherry SR, Barrio JR, Satyamurthy N, Toyokuni T, Phelps ME, Larson SM, Balatoni J, Finn R, Sadelain M, Tjuvajev J, Blasberg R. 2000. Imaging transgene expression with radionuclide imaging technologies. Neoplasia 2:118-138.
Graichen H, Lochmuller EM, Wolf E, Langkabel B, Stammberger T, Haubner M, Renner-Muller I, Englmeier KH, Eckstein F. 1998. A non-destructive technique for 3-D microstructural phenotypic characterisation of bones in genetically altered mice: preliminary data in growth hormone transgenic animals and normal controls. Anat Embryol (Berl) 199:239-248.
Hassel B, Brathe A. 2000. Cerebral metabolism of lactate in vivo: evidence for neuronal pyruvate carboxylation. J Cerebr Blood Flow Metab 20:327-336.
Hedlund LW, Cofer GP, Owen SJ, Allan JG. 2000. MR-compatible ventilator for small animals: computer-controlled ventilation for proton and noble gas imaging. Magn Reson Imaging 18:753-759.
Hees PS, Sotak CH. 1993. Assessment of changes in murine tumor oxygenation in response to nicotinamide using 19F NMR relaxometry of a perfluorocarbon emulsion (published erratum appears in Magn Reson Med 1993 29:716). Magn Reson Med 29:303-310.
Holtzman D, Meyers R, O'Gorman E, Khait I, Wallimann T, Allred E, Jensen F. 1997. In vivo brain phosphocreatine and ATP regulation in mice fed a creatine analog. Am J Physiol 272(Pt 1):C1567-C1577.
Hsieh P, Balaban RS. 1987. 31P Imaging of in vivo creatine kinase reaction rates. J Magn Reson 74:574-579.
Huang WY, Aramburu J, Douglas PS, Izumo S. 2000. Transgenic expression of green fluorescence protein can cause dilated cardiomyopathy. Nat Med 6:482-483.
Hurlston SE, Brey WW, Suddarth SA, Johnson GA. 1999. A high-temperature superconducting Helmholtz probe for microscopy at 9.4 T. Magn Reson Med 41:1032-1038.
Ikehira H, Yamane T, Fukuda N, Ando K, Aoki Y, Koike S, Endo M, Matumoto T, Iinuma T, Fukuda H. 1988. Fundamental tumor perfusion analysis with nuclear magnetic resonance imaging using gadolinium-diethylene triamine pentaacetic acid. Am J Physiol Imaging 3:7-9.
in 't Zandt HJ, Oerlemans F, Wieringa B, Heerschap A. 1999. Effects of ischemia on skeletal muscle energy metabolism in mice lacking creatine kinase monitored by in vivo 31P nuclear magnetic resonance spectroscopy. NMR Biomed 12:327-334.
Ingvar M, Eriksson L, Rogers GA, Stone-Elander S, Widen L. 1991. Rapid feasibility studies of tracers for positron emission tomography: high-resolution PET in small animals with kinetic analysis. J Cerebr Blood Flow Metab 11:926-931.
Jacobs RE, Ahrens ET, Dickinson ME, Laidlaw D. 1999. Towards a microMRI atlas of mouse development. Comput Med Imaging Graph 23:15-24.
Kanamori K, Ross BD. 1997. Glial alkalinization detected in vivo by 1H-15N heteronuclear multiple-quantum coherence-transfer NMR in severely hyperammonemic rat. J Neurochem 68:1209-1220.
Kehlet BJ, Izatt JA, Kulkarni MD, Yazdanfar S, Welch AJ. 1999. Three-dimensional reconstruction of blood vessels from in vivo color Doppler optical coherence tomography images. Dermatology 198:355-361.
Kennel SJ, Davis IA, Branning J, Pan H, Kabalka GW, Paulus MJ. 2000. High resolution computed tomography and MRI for monitoring lung tumor growth in mice undergoing radioimmunotherapy: Correlation with histology. Med Phys 27:1101-1107.
Knuttel A, Boehlau-Godau M. 2000. Spatially confined and temporally resolved refractive index and scattering evaluation in human skin performed with optical coherence tomography. J Biomed Opt 5:83-92.
Kuge Y, Kawashima H, Hashimoto T, Imanishi M, Shiomi M, Minematsu K, Hasegawa Y, Yamaguchi T, Miyake Y, Hashimoto N. 2000. Preliminary evaluation of [1-11C]octanoate as a PET tracer for studying cerebral ischemia: A PET study in rat and canine models of focal cerebral ischemia. Ann Nucl Med 14:69-74.
Lauffer RB, Parmelee DJ, Dunham SU, Ouellet HS, Dolan RP, Witte S, McMurry TJ, Walovitch RC. 1998. MS-325: albumin-targeted contrast agent for MR angiography. Radiology 207:529-538.
Lendvai B, Stern EA, Chen B, Svoboda K. 2000. Experience-dependent plasticity of dendritic spines in the developing rat barrel cortex in vivo. Nature 404:876-881.
Levchenko A, Mehta BM, Lee JB, Humm JL, Augensen F, Squire O, Kothari PJ, Finn RD, Leonard EF, Larson SM. 2000. Evaluation of 11C-colchicine for PET imaging of multiple drug resistance. J Nucl Med 41:493-501.
Linask KK, Huhta JC. 2000. Use of Doppler echocardiography to monitor embryonic mouse heart function. Methods Mol Biol 135:245-252.
Lindner JR, Dayton PA, Coggins MP, Ley K, Song J, Ferrara K, Kaul S. 2000. Noninvasive imaging of inflammation by ultrasound detection of phagocytosed microbubbles. Circulation 102:531-538.
Linney AD, Deng J. 1999. Three-dimensional morphometry in ultrasound. Proc Inst Mech Eng 213:235-245.
Lukes D, Madhou B, Arvidsson N, Gustavsson M, Mjornstedt L, Soussi B, Olausson M. 1997. In vivo 31P MRS evaluation of the rejection process and differences in anesthetic procedures in a concordant xenotransplantation: mouse heart to rat modell. Transplant Proc 29:3159-3160.
Lyng H, Dahle GA, Kaalhus O, Skretting A, Rofstad EK. 1998. Measurement of perfusion rate in human melanoma xenografts by contrast-enhanced magnetic resonance imaging. Magn Reson Med 40:89-98.
Magata Y, Saji H, Choi SR, Tajima K, Takagaki T, Sasayama S, Yonekura Y, Kitano H, Watanabe M, Okada H. 1995. Noninvasive measurement of cerebral blood flow and glucose metabolic rate in the rat with high-resolution animal positron emission tomography (PET): A novel in vivo approach for assessing drug action in the brains of small animals. Biol Pharm Bull 18:753-756.
Mason RP, Shukla H, Antich PP. 1993. In vivo oxygen tension and temperature: simultaneous determination using 19F NMR spectroscopy of perfluorocarbon. Magn Reson Med 29:296-302.
Masters BR, So PT, Gratton E. 1998. Multiphoton excitation microscopy of in vivo human skin. Functional and morphological optical biopsy based on three-dimensional imaging, lifetime measurements and fluorescence spectroscopy. Ann NY Acad Sci 838:58-67.
Maurer J, Strauss A, Ebert W, Bauer H, Felix R. 2000. Contrast-enhanced high resolution magnetic resonance imaging of pigmented malignant melanoma using Mn-TPPS4 and Gd-DTPA: experimental results. Melanoma Res 10:40-46.
Maxwell RJ, Nielsen FU, Breidahl T, Stodkilde-Jorgensen H, Horsman MR. 1998. Effects of combretastatin on murine tumours monitored by 31P MRS, 1H MRS and 1H MRI. Int J Radiat Oncol Biol Phys 42:891-894.
McSheehy PM, Seymour MT, Ojugo AS, Rodrigues LM, Leach MO, Judson IR, Griffiths JR. 1997. A pharmacokinetic and pharmacodynamic study in vivo of human HT29 tumours using 19F and 31P magnetic resonance spectroscopy. Eur J Cancer 33:2418-2427.
Meynial-Denis D, Chavaroux A, Foucat L, Mignon M, Prugnaud J, Bayle G, Renou JP, Arnal M. 1997. Contribution of proteolysis and de novo synthesis to alanine production in diabetic rat skeletal muscle: A 15N/1H nuclear magnetic resonance study. Diabetologia 40:1159-1165.
Miller JR, Hurlston SE, Ma QY, Face DW, Kountz DJ, MacFall JR, Hedlund LW, Johnson GA. 1999. Performance of a high-temperature superconducting probe for in vivo microscopy at 2.0 T. Magn Reson Med 41:72-79.
Minard KR, Wind RA, Phelps RL. 1998. A compact respiratory-triggering device for routine microimaging of laboratory mice. J Magn Reson Imaging 8:1343-1348.
Nielsen FU, Topp S, Horsman MR, Overgaard J, Stodkilde-Jorgensen H, Maxwell RJ. 1999. Localized in vivo 1H NMR spectroscopy of murine tumours: Effect of blood flow reduction. NMR Biomed 12:175-183.
Pascual M, Jahoor F, Reeds PJ. 1998. In vivo glucose contribution to glutamate synthesis is maintained while its contribution to acetyl CoA is lowered in adult mice fed a diet with a high fat:carbohydrate ratio. J Nutr 128:733-739.
Paulus MJ, Gleason SS, Kennel SJ, Hunsicker PR, Johnson DK. 2000. High resolution x-ray computed tomography: an emerging tool for small animal cancer research. Neoplasia 2:62-70.
Pautler RG, Silva AC, Koretsky AP. 1998. In vivo neuronal tract tracing using manganese-enhanced magnetic resonance imaging. Magn Reson Med 40:740-748.
Peled-Kamar M, Degani H, Bendel P, Margalit R, Groner Y. 1998. Altered brain glucose metabolism in transgenic-PFKL mice with elevated L-phosphofructokinase: in vivo NMR studies. Brain Res 810:138-145.
Phelps ME. 2000. Inaugural article: Positron emission tomography provides molecular imaging of biological processes. Proc Natl Acad Sci U S A 97:9226-9233.
Robinson SP, van den BA, Maxwell RJ, Griffiths JR, Hamilton E, Waterton JC. 1998. 31P-magnetic resonance spectroscopy and 2H-magnetic resonance imaging studies of a panel of early-generation transplanted murine tumour models. Br J Cancer 77:1752-1760.
Ronen I, Merkle H, Ugurbil K, Navon G. 1998. Imaging of H217O distribution in the brain of a live rat by using proton-detected 17O MRI. Proc Natl Acad Sci U S A 95:12934-12939.
Roper SN, Moores MD, Gelikonov GV, Feldchtein FI, Beach NM, King MA, Gelikonov VM, Sergeev AM, Reitze DH. 1998. In vivo detection of experimentally induced cortical dysgenesis in the adult rat neocortex using optical coherence tomography. J Neurosci Methods 80:91-98.
Ros PR, Li KC, Vo P, Baer H, Staab EV. 1990. Preautopsy magnetic resonance imaging: initial experience. Magn Reson Imaging 8:303-308.
Saini SK, Jena A, Dey J, Sharma AK, Singh R. 1995. MnPcS4: A new MRI contrast enhancing agent for tumor localisation in mice. Magn Reson Imaging 13:985-990.
Schornack PA, Song SK, Ling CS, Hotchkiss R, Ackerman JJ. 1997. Quantification of ion transport in perfused rat heart: 133Cs+ as an NMR active K+ analog. Am J Physiol 272(Pt 1):C1618-C1634.
Sebrie, Gillet B, Lefaucheur JP, Sebille A, Beloeil JC. 1998. Mouse muscle regeneration: An in vivo 2D 1H magnetic resonance spectroscopy (MRS) study. FEBS Lett 423:71-74.
Sibson NR, Shen J, Mason GF, Rothman DL, Behar KL, Shulman RG. 1998. Functional energy metabolism: In vivo 13C-NMR spectroscopy evidence for coupling of cerebral glucose consumption and glutamatergic neuronalactivity. Dev Neurosci 20:321-330.
Slawson SE, Roman BB, Williams DS, Koretsky AP. 1998. Cardiac MRI of the normal and hypertrophied mouse heart. Magn Reson Med 39:980-987.
Smith BR. 2000. Magnetic resonance imaging analysis of embryos. Methods Mol Biol 135:211-216.
Sonner JM, Gong D, Eger EI. 2000. Naturally occurring variability in anesthetic potency among inbred mouse strains. Anesth Analg 91:720-726.
Sonner JM, Gong D, Li J, Eger EI, Laster MJ. 1999. Mouse strain modestly influences minimum alveolar anesthetic concentration and convulsivity of inhaled compounds. Anesth Analg 89:1030-1034.
Squirrell JM, Wokosin DL, White JG, Bavister BD. 1999. Long-term two-photon fluorescence imaging of mammalian embryos without compromising viability. Nat Biotechnol 17:763-767.
Steen RG, Wilson DA, Bowser C, Wehrle JP, Glickson JD, Rajan SS. 1989. 31P NMR spectroscopic and near infrared spectrophotometric studies of effects of anesthetics on in vivo RIF-1 tumors. Relationship to tumor radiosensitivity. NMR Biomed 2:87-92.
Steinberg F, Rohrborn HJ, Otto T, Scheufler KM, Streffer C. 1997. NIR reflection measurements of hemoglobin and cytochrome aa3 in healthy tissue and tumors. Correlations to oxygen consumption: Preclinical and clinical data. Adv Exp Med Biol 428:69-77.
Syme PD, Dixon RM, Aronson JK, Grahame-Smith DG, Radda GK. 1990. Evidence for increased in vivo sodium-potassium pump activity and potassium efflux in skeletal muscle of spontaneously hypertensive rats. J Hypertens 8:1161-1166.
Takeishi Y, Ping P, Bolli R, Kirkpatrick DL, Hoit BD, Walsh RA. 2000. Transgenic overexpression of constitutively active protein kinase C epsilon causes concentric cardiac hypertrophy. Circ Res 86:1218-1223.
Turnbull DH. 2000. Ultrasound backscatter microscopy of mouse embryos. Methods Mol Biol 135:235-243.
Ueda H, Hashimoto T, Furuya E, Tagawa K, Kitagawa K, Matsumoto M, Yoneda S, Kimura K, Kamada T. 1988. Changes in aerobic and anaerobic ATP-synthesizing activities in hypoxic mouse brain. J Biochem (Tokyo) 104:81-86.
Viallon M, Cofer GP, Suddarth SA, Moller HE, Chen XJ, Chawla MS, Hedlund LW, Cremillieux Y, Johnson GA. 1999. Functional MR microscopy of the lung using hyperpolarized 3He. Magn Reson Med 41:787-792.
Villanueva FS, Jankowski RJ, Klibanov S, Pina ML, Alber SM, Watkins SC, Brandenburger GH, Wagner WR. 1998. Microbubbles targeted to intercellular adhesion molecule-1 bind to activated coronary artery endothelial cells. Circulation 98:1-5.
Weissleder R. 1999. Molecular imaging: exploring the next frontier. Radiology 212:609-614.
Weissleder R, Moore A, Mahmood U, Bhorade R, Benveniste H, Chiocca EA, Basilion JP. 2000. In vivo magnetic resonance imaging of transgene expression. Nat Med 6:351-355.
Wiesmann F, Ruff J, Haase A. 1998. High-resolution MR imaging in mice. MAGMA 6:186-188.
Winter PM, Seshan V, Makos JD, Sherry AD, Malloy CR, Bansal N. 1998. Quantitation of intracellular (Na+) in vivo by using TmDOTP5- as an NMR shift reagent and extracellular marker. J Appl Physiol 85:1806-1812.
Wolff SD, Eng J, Berkowitz BA, James S, Balaban RS. 1990. Sodium-23 nuclear magnetic resonance imaging of the rabbit kidney in vivo. Am J Physiol 258(Pt 2):F1125-F1131.
Xia ZF, Horton JW, Zhao PY, Babcock EE, Sherry AD, Malloy CR. 1996. Effects of ischemia on intracellular sodium and phosphates in the in vivo rat liver. J Appl Physiol 81:1395-1403.
Yamamoto S, Teng W, Nishizawa S, Kakiuchi T, Tsukada H. 2000. Improvement in cerebral blood flow and metabolism following subarachnoid hemorrhage in response to prophylactic administration of the hydroxyl radical scavenger, AVS, (+/-)-N,N'-propylenedinicotinamide: A positron emission tomography study in rats. J Neurosurg 92:1009-1015.
Yamashita T, Nabeshima Y, Noda M. 2000. High-resolution micro-computed tomography analyses of the abnormal trabecular bone structures in klotho gene mutant mice. J Endocrinol 164:239-245.
Yokosawa K, Sasaki K, Umemura S, Shinomura R, Ishikawa S, Sano S, Ito Y. 2000. Intracorporeal imaging and differentiation of living tissue with an ultra-high-frequency ultrasound probe. Ultrasound Med Biol 26:503-507.
Yu Y, Annala AJ, Barrio JR, Toyokuni T, Satyamurthy N, Namavari M, Cherry SR, Phelps ME, Herschman HR, Gambhir SS. 2000. Quantification of target gene expression by imaging reporter gene expression in living animals. Nat Med 6:933-937.
Zimmermann U, Noth U, Grohn P, Jork A, Ulrichs K, Lutz J, Haase A. 2000. Non-invasive evaluation of the location, the functional integrity and the oxygen supply of implants: 19F nuclear magnetic resonance imaging of perfluorocarbon-loaded Ba2+-alginate beads. Artif Cells Blood Substit Immobil Biotechnol 28:129-146.
Figure 1 Schematic diagram of information content of magnetic resonance imaging (MRI)/magnetic resonance spectroscopy (MRS). Italicized text is information extracted from water proton imaging (MRI). Most of the other information is gathered using MRS techniques. ATP, adenosine 5c-triphosphate; GABA, aminobutyric acid; GPC, glycerolphosphorylcholine; GPE, glycerolphosphorylethanolamine; NAA, N-acetylaspartate; NAD, nicotinamide adenine dinucleotide; PC, phosphorylcholine; PE, phosphorylethanolamine.
Figure 2 Proton magnetic resonance images of the mouse (left) and human heart (right). The mouse heart image was collected in a 4.7-T system using a surface coil placed around the chest of the mouse. The human image was collected at 1.5 T using a whole body coil and a similar inversion recovery sequence as used in the mouse to result in the "black-blood" image. Both images were collected in less than 5 min.
Figure 3 45-MHz ultrasound images of the mouse heart. Data used with permission from Fatkin D, Christe ME, Aristizabal O, McConnell BK, Srinivasan S, Schoen FJ, Seidman CE, Turnbull DH, Seidman JG. 1999. Neonatal cardiomyopathy in mice homozygous for the Arg403Gln mutation in the alpha cardiac myosin heavy chain gene. J Clin Invest 103:147-153.
Figure 4 Computer-assisted tomography coronal section through a whole mouse. Image courtesy of Oak Ridge National Laboratories, Oak Ridge, Tennessee.
Figure 5 18-F positron emission tomography (PET) images of a mouse after virus infection with the HSV1-tk reporter gene. 8-(18F) fluoroganciclovir was used as the active metabolite. (a) Control images. (b) Active virus. The labeling in the liver and intestine was found to be enhanced in the presence of the viral infection. ID/g, infectious dose per gram. Data used with permission from Gambhir SS, Herschman HR, Cherry SR, Barrio JR, Satyamurthy N, Toyokuni T, Phelps ME, Larson SM, Balatoni J, Finn R, et al. 2000. Imaging transgene expression with radionuclide imaging technologies. Neoplasia 2:118-138.
Figure 6 (a) Green fluorescence confocal microscopy of dendrites in developing rat brain. (b) Images of barrel cortex neurons that were enhanced by the expression of green fluorescence protein DNA introduced by the selective injection of engineered Sindbis virus. Data used with permission from Lendvai B, Stern EA, Chen B, Svoboda K. 2000. Experience-dependent plasticity of dendritic spines in the developing rat barrel cortex in vivo. Nature 404:876-881.
Figure 7 Whole body near infrared (IR) fluorescence image of a mouse. (A) Mouse reference image. (B) Whole body infrared fluorescence. A polymer containing a self-quenched fluorescent probe was administered by injection into the animal and trapped in the tumor (bright structure in infrared image). The fluorescence was enhanced in the tumor by the activity of lysosomal proteases, which cleaved the polymer and reduced the self-quenching of the probe by releasing it from the polymer. Data used with permission from Weissleder R. 1999. Molecular imaging: Exploring the next frontier. Radiology 212:609-614.
Table 1 Justifications of vital imaging studies
· Need for dynamic physiological data or labile biochemical/structural information
· Need for internal longitudinal controls
· Animals are valuable breeding stock or difficult to produce or maintain
· Monitoring and adjustments of drug treatments or other experimental protocols
· Use of imaging technique to screen for inclusion in another protocol
Copyright © 2008. National Academy of Sciences.
All rights reserved.
500 Fifth St. N.W., Washington, D.C. 20001.
Terms of Use and Privacy Statement