Technical Background

The development of a detailed understanding of the physics principles that underlie MRS is not achievable in a short chapter such as this. However, the following paragraphs provide a succinct, accurate but somewhat superficial description of important principles and concepts.

MRS is closely related to nuclear magnetic resonance (NMR) spectroscopy, which has been used in chemistry and physics since the late 1940s. The term "MRS" is now commonly used for biomedical applications, in preference to "NMR" or "NMR spectroscopy,'' when NMR spectroscopy studies of living human or animal subjects or cultured living cells are undertaken. MRS is viewed by the radiology and medical imaging communities as member of a large family of MRI techniques. MRS differs from MRI in that MRS detects signals from chemical compounds other than water to evaluate in situ biochemistry, whereas MRI detects tissue water or lipid signals to form images that depict anatomy. However, even this distinction is somewhat simplistic; MRS techniques that image the spatial distribution of in situ biochemicals have been used for the evaluation of cancers since the early 1990s (14,15). MRS and MRI are each performed with an MRI scanner. There are thousands of MRI scanners installed in academic radiology departments and in some private radiology practices throughout the world that are capable of performing MRS. However some MRI scanners are not suitable for MRS. Typically, MRS requires a static magnetic field strength of greater than 1.0 T. Low-field MRI scanners can not therefore be used. As is the case with MRI, MRS requires only exposure to time invariant (static) and oscillatory magnetic fields. Hence, one of the greatest attributes of MRS is that it can evaluate in situ biochemistry without exposing the subject to ionizing radiation or to radioactive isotopes.

Discussion of the component terms of MRS (i.e. magnetic, resonance, and spectroscopy) provides a slightly more detailed understanding. The term "magnetic"

is used because MRS utilizes several special types of magnetic fields. Certain atomic nuclei behave as if they were spinning and this causes such nuclei to appear as if they were minute bar magnets. This "nuclear magnetism'' can interact with a number of external magnetic fields including (i) the static magnetic field created by the MRI magnet, (ii) the oscillating magnetic fields produced by the MRI scanner's radiofre-quency (RF) coils, and (iii) the pulsed magnetic field distortions produced by the MRI scanner's magnetic field gradient system. These ''magnetic interactions" lead to a particular type of energy exchange, referred to as "resonance," between the scanner and the nuclear magnetic fields and it is this exchange that is detected as the "MRS signal.'' The term "spectroscopy" conveys the idea that the magnetic resonance occurs only at specific frequencies and that the signal-producing nuclei can be identified by the presence of resonance only at specific characteristic frequencies. For clinical MRS studies, the characteristic resonance frequency has a time dependence that is similar to that of the "radio waves'' used in radio and television broadcasting (Fig. 1). Hence MRS is sometimes described as using radio waves. The basic MRS result is the ''spectrum,'' which is a two-dimensional plot of frequency on the horizontal axis and intensity of resonance interaction on the vertical axis (Fig. 1).

Only certain atomic nuclei (isotopes) of biological significance (e.g., 1H,31P, 13C,7Li, and 19F) have suitable magnetic properties and are therefore capable of producing MRS signals (Fig. 1). MRS signals produced by these isotopes are easily distinguished from each other by their much different characteristic frequencies. The ability of MRS to detect unique signals from different chemicals results from the fact that the magnetic resonance frequency is directly proportional to the static magnetic field strength at the nucleus. The electrons which surround the atomic nuclei located within molecules circulate in ways that tend to alter the magnetic field at the atomic nucleus to a small but significant extent. This causes a small but detectable alterations in the MRS signal frequency which are dependent on the chemical structure surrounding an atomic nucleus, allowing the identification of specific resonance signals from individual nuclei within individual molecules (Figs. 1 and 2). The differences in frequency resulting from chemical structure are referred to as ''chemical shifts,'' which are frequently rather small. Important MRS signals from a particular isotope, such as 1H, may be separated from each other in frequency by only a few cycles per second [Hertz (Hz)] with each of their signals having a frequency of millions of cycles per second [Megahertz (MHz)]. For this reason, it is a common practice to specify the chemical shift in ''parts per million (ppm).''

The different nuclei of biological significance have different attributes and practical limitations with respect to MRS detection. The proton [1H] produces the strongest most easily detected MRS signal and is therefore most frequently used for routine clinical MRS. Moreover,1H-MRS is convenient in that it can be performed using the same hardware as is used for conventional MRI. 31P produces the second most intense MRS signal. Figures 1 and 2 illustrate general features that are seen in 1H- and 31P-MRS of normal and neoplastic brain tissue. 31P-MRS has been the basis for some clinical MRS examinations. Indeed, many of the early MRS studies of cancer were performed with 31P-MRS, but the difficulty associated with detection of 31P signals compared to 1H-MRS signals has led to far more common use of 1H-MRS in recent years. Other atomic nuclei are of research interest, but are not used in routine clinical MRS studies of cancer. For this reason, this chapter will discuss only 1H- and 31P-MRS.

Figure 1 Characteristic MRS signal frequencies. Different stable nuclear isotopes of biological significance produce MRS signals at unique and characteristic frequencies. The characteristic nuclear frequencies differ by a great amount and are easily distinguished. For instance, all nuclei produce signals proximal to 63 MHz in a magnetic field of 1.5 T, whereas all 31P nuclei produce signal proximal to 26 MHz when the same magnetic field strength is used. Within a narrow band of frequencies surrounding the characteristic signal frequency of each nucleus are found distinct signals from chemically unique nuclei. Typical 1H and 31P spectra of normal living brain tissue are shown in the insets. The 1H spectrum shows a large signal produced by tissue water that frequently needs to be suppressed to detect the smaller signals from biochemicals within the tissue. Assignments for chemically unique signals relevant to cancer are given in Figure 2. The 31P spectrum shows signals from NTP and other tissue metabolities associated with energy metabolism. The phosphomonoester signal is the most relevant for cancer studies because many cancers show an elevation of this signal. In addition, the frequency difference between the phosphocreatine and inorganic phosphate signals is sensitive to intracellular pH permitting the noninvasive determination of pH within cancer cells tumors. Abbreviations: NTP, nucleotide triphosphate; MHz, megahertz; T, Tesla; ppm, parts per million.

Clinical MRS research studies involving cancer have most frequently sought to evaluate neoplastic mass lesions. In order to evaluate a mass lesion with MRS, there is a need for procedures that can localize the anatomic source of the MRS signals that are detected and studied. Figures 2 and 3 provide illustrations of the two complementary methodologies that are available for attaining volume localization. In localized single volume MRS (Fig. 2), a conventional MR image is used to identify a location of interest within or adjacent to the tumor which is typically defined as a rectilinear "voxel," and MRS signal is acquired from only this location (5,16). If it is desired to obtain MRS from other locations, the localized single volume MRS procedure is repeated. In magnetic resonance spectroscopic imaging (MRSI),

Figure 2 Single volume localized 1H-MRS data from normal human brain tissue (bottom panel) and from a contrast enhancing brain tumor (top panel). When using the single volume localized MRS technique, a region of interest is defined using MRI. Without moving the patient, 1H-MRS data are then collected from the region of interest using a localized spectroscopy pulse sequence. Regions of interest for the two studies on the MRIs are shown (white rectangles). The resultant spectra are shown to the right of the MRIs. Key 1H-MRS signals are labeled. Note that the choline signal is substantially more elevated above baseline in the tumor spectrum compared to the normal spectrum. Furthermore, there is a lactate signal present in the tumor spectrum but not in the normal spectrum and the NAA signal is reduced in the tumor spectrum compared to the normal spectrum. Note that some tumors can also produce strong lipid signals which can often not be distinguished from lactate signals, although this is not the case in this example (44). There is, in fact, a great deal of variability in the spectroscopic patterns exhibited by different tumors at different stages of treatment and response. Moreover there is even variation of spectroscopic patterns displayed by normal brain tissue. Abbreviations: NAA, N-acetylaspartate; 1H-MRSI, 1H-magnetic resonance spectroscopic imaging; ppm, parts per million.

MRS signals are simultaneously acquired from a grid containing a large number of rectilinear voxels that include the tumor and surrounding tissues that is prescribed from a preliminary MRI study (Fig. 3) (8,14,15,17). MRSI is also sometimes referred to as "spectroscopic imaging (SI)'' or "chemical shift imaging (CSI)." Figure 3 illustrates that MRSI is capable of providing either localized spectra of individual voxels that can be chosen in a post hoc fashion, or "SI" of the anatomic distribution of the intensity of a particular signal.

Figure 3 Typical 1H-MRSI results from a glioblastoma multiforme. (A) When using the 1H-MRSI technique, a slice or volume of tissue of interest is defined from MRI. (B) Spectroscopic imaging procedures are then used to obtain a spectrum from each of the locations shown by the grid. Selected spectra sampled from this grid are shown for (C) tumor tissue and (D) nearby normal tissue. The area under relevant signals is then determined in each of the spectra over the entire grid. These signal measures are then represented with color spectroscopic images [(E) the choline signal image and (F) the NAA signal image]. Note that the spectra [(B) and (C)] and spectroscopic images [(E) and (F)] show NAA signal decrease and choline signal increase in the tumor compared to normal tissue. As with Figure 2, it is important to emphasize that these patterns are characteristic of tumor and normal tissue, but that each type of tissue displays considerable variability. Abbreviations'. 1H-MRSI, 1H-magnetic resonance spectroscopic imaging; NAA, N-acetylaspartate; ppm, parts per million.

Figure 3 Typical 1H-MRSI results from a glioblastoma multiforme. (A) When using the 1H-MRSI technique, a slice or volume of tissue of interest is defined from MRI. (B) Spectroscopic imaging procedures are then used to obtain a spectrum from each of the locations shown by the grid. Selected spectra sampled from this grid are shown for (C) tumor tissue and (D) nearby normal tissue. The area under relevant signals is then determined in each of the spectra over the entire grid. These signal measures are then represented with color spectroscopic images [(E) the choline signal image and (F) the NAA signal image]. Note that the spectra [(B) and (C)] and spectroscopic images [(E) and (F)] show NAA signal decrease and choline signal increase in the tumor compared to normal tissue. As with Figure 2, it is important to emphasize that these patterns are characteristic of tumor and normal tissue, but that each type of tissue displays considerable variability. Abbreviations'. 1H-MRSI, 1H-magnetic resonance spectroscopic imaging; NAA, N-acetylaspartate; ppm, parts per million.

10 Ways To Fight Off Cancer

10 Ways To Fight Off Cancer

Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.

Get My Free Ebook


Post a comment