Figure 14.2 An MR spectrum reveals information on the chemical identity (position on the horizontal axis) and the concentration (area under the resonance) of metabolites. The example shows three different chemical groups within butyrate, each group with a specific position on the chemical shift axis. The splitting of the resonance lines is an effect of the number of nearest neighbours in a molecule and can additionally be used to identify a resonance spectrum defines the chemical nature while the signal area reveals the amount of molecule. The width of a resonance line is another important experimental factor since it defines the resolution of the spectrum, i.e. how many chemical species can be observed separately. In addition to the type of chemical compound, the homogeneity of the magnetic field influences the width of the lines, i.e. the resolution of a spectrum. The homogeneity of the magnetic field can be improved by technical means, called 'shimming', in turn biological tissue can introduce additional inhomogeneity of the field, particularly at air-tissue borders. As we will see below, gas bubbles that are produced during the process of body decomposition can severely decrease the homogeneity of the magnetic field such that spectral resolution becomes a problem.
While high-resolution NMR uses small sample tubes in vertical magnets up to a magnetic field strength of 22 Tesla, in situ and in vivo MR spectra are acquired in horizontal bore magnets with typical field strengths of 1.5 or 3 Tesla. Selection of the signal-generating tissue ('voxel', 'region of interest ROI') is achieved by techniques that are common with MRI. The voxel can be placed at the desired anatomical location based on a series of localizer MR images (see Figure 14.3 opposite ) and contains typically a few millilitres in 1H-MRS. An MR spectrum can be acquired by technicians in much less than an hour without the need for sample preparation and a lH-MR-spectrum contains information on about 20 metabolites simultaneously. Depending on the automation of the spectral analysis, final data can be obtained in less than an hour.
In order to estimate the PMI, 1H-MRS of the brain is particularly advantageous:
1. 1H-MRS provides high relative sensitivity, allowing the selection of small voxels.
2. Established MRS sequences for 1H allow for a robust and easy selection of spectroscopic volumes with subsequent absolute quantitation of the metabolites.
3. A majority of clinical MRS investigations are performed on the brain, resulting in widespread experimental experience and a large collection of reference data in healthy and pathological brain tissue.
4. About 20 different characterized substances can be observed and quantified.
5. Inter-individual differences in brain tissue composition are rather small (approx. 5%; Kreis, 1997) and can partly be attributed to the aging process (Chang et al., 1996), which can be corrected for.
Figure 14.3 Magnetic resonance images of a sheep brain with a placement of voxels for the spectroscopic data acquisition. The images are so called T2-weighted, i.e. the signal acquisition is optimized such that the contrast between different types of tissue is optimal. During decomposition of the brain, voxel placement has to avoid gas bubbles
6. The brain is protected by the skull after death, i.e. destruction by environmental factors (scavengers, external microorganisms) is minimized, and when the brain tissue becomes decomposed and liquefied the skull guarantees a certain fixation.
A determination of late PMIs by means of 1H-MRS of the brain is based on the fact that MRS reveals concentrations of multiple metabolites in a single spectrum. If there are calibration curves of metabolite changes following death, it should be possible to compare the measured concentrations in a brain with an unknown time of death with the calibration curves. The following paragraphs will illustrate how calibration curves are established in an animal model, how first cases of human bodies are studied and which statistical procedures are used to compare actual metabolite concentrations and calibration curves.
Ethical reasons prevent the storage of human bodies longer than necessary for a regular forensic examination, making it almost impossible to establish calibration curves of metabolite changes over a longer time based on human bodies. An animal model allows decomposition processes of the brain to be followed repeatedly and under standardized conditions at different, optimally spaced points in time.
While in the following study a sheep model is used, there have also been reports on studies in pigs (Banaschak et al., 2005). For practical reasons, the animal has to be a mammal and should be available in a slaughterhouse. The size of the animal brain should be as large as possible to optimize voxel selection; however, the complete head should still be within the size of an MR head-coil. As described in detail by Ith et al. (2002), the heads used in the following study were harvested from healthy sheep, which died in the course of the normal slaughtering process. The heads were separated from the bodies and stored at constant temperature (21 ± 3°C). In situ brain spectra from the frontal lobe and the parieto-occipital region (Figure 14.3) were acquired regularly up to 18 days postmortem. A striking increase of the signal amplitude shows how decomposition of brain tissue makes many more metabolites MR-visible (Figure 14.4). Figure 14.4 also illustrates that other chemical compounds are formed during decomposition.
Figure 14.4 Two spectra of sheep brain at the same scale, the lower trace recorded in vivo and the upper trace 18 days postmortem. Following death, new metabolites appear and the concentration of visible metabolites increases dramatically in the decomposing brain tissue
In order to obtain quantitative data, absolute concentrations of the metabolites had to be determined. This is achieved by a comparison of the signals with the signal from water (fully relaxed) from the same volume and with an estimation of the resonance area by a mathematical modelling of the experimental curve, i.e. by a fitting procedure. A well-known fitting algorithm is 'LC Model' (Provencher, 1993) that uses linear combinations of model-spectra, a so-called 'basis set' of spectra from measurable substances. The basis set, which in clinical practice is used to fit brain spectra, consists of about 25 metabolites: acetate, alanine, aspartate, cholines, creatines, phospho-ethanolamine, ethanol, gamma-aminobutyric acid, glucose, glutamine, glutamate, glutathione, glycine, lactate, myo-inositol, scyllo-inositol, N-acetyl-asparates, taurine, and valine. However, during the particularly interesting late postmortem phase, several substances appeared that had not been observed in brain spectra before then, and had to be identified by means of high-resolution NMR as trimethylamine, propionate, butyrate and isobutyrate (Ith et al., 2002). Since the additional metabolites are detected only after the third day postmortem and do not contribute significantly to spectra before then, they are most likely of bacterial origin. A chemical identification of substances is not only necessary for a deeper understanding of the underlying processes, but also to analyse the spectra mathematically. An incomplete set of theoretical spectra results in a poorer fit, and spectral features could be wrongly attributed to substances that are included in the basis set. This would lead to an underestimation of the unknown substance and, subsequently, to an erroneous overestimation of an included substance with similar spectral features. Therefore, a complete set of spectra improves the accuracy for all substances in a spectrum. Inclusion of the newly identified metabolites and succinate (corrected for pH and temperature effects) resulted in a significantly improved fit quality and reliable quantitation of up to 30 metabolites.
The time course of the observed metabolites has then been followed up to 18 days postmortem. While some of the time courses are ambiguous or too scattered, others show a clear and unequivocal time dependence, such as NAA/ NAAG or butyrate as shown in Figure 14.5. Obviously, the decomposition process follows certain rules that can be used for an estimation of the PMI. In the early postmortem phase metabolic changes probably due to autolytic processes are observed, i.e., a decay of the 'neuronal marker' N-acetyl-aspartate (Figure 14.5a) into acetate and aspartate and the build-up of lactate during glycolysis. Later, starting at about 3 days postmortem, bacterial decomposition processes, i.e. heterolysis, come up in addition to autolysis, leading to an observation of substances that are associated with the action of bacteria, e.g. butyrate (Figure 14.5b), trimethylamine or propionate.
Was this article helpful?
When it comes to reducing wrinkles, you really have to take your needs seriously. There are a number of factors that play a role in the health and well being of your skin. It is often hard to understand how products work and why they may not work even if they promise to do so. If you are considering wrinkle creams and a regimen of facelifts, you may go broke in the process. Before you do all of that, consider going through a process of determine the very best solution for your needs.