This research focused on demonstrating the simultaneous collection of optical and NMR spectra from in vivo biological samples and the engineering challenges to be met in accomplishing this goal. The direct results were the design and fabrication of a fiber-optically coupled CCD based OMA system capable of recording fluorescence signals from a sample undergoing NMR interrogation in the bore of a 2 T superconducting magnet and the design of an NMR probe capable of collecting ³¹P NMR spectra from the in situ, in vivo heart of the western painted turtle Chysemys picta bellii.

7.1 Summary

An optical system capable of collecting small intensity optical fluorescence signals from a remote sample in the bore of a high field NMR magnet was designed, constructed, and tested. The instrument used fiber-optics to transmit light to the sample and back to the instrument. A thermoelectrically cooled CCD array and imaging spectrograph were used to construct a sensitive OMA for simultaneous recording of the entire optical emission spectrum from the sample.

The use of an array of microcoils was explored as a means to increase SNR and tailor the field of view for NMR measurements from a specific region of interest. A physiological model of the turtle heart and blood plasma was formulated and an electromagnetic modeling program was used to investigate the relative SNR of different sized array elements. It was found that for the particular sample of interest, an array of small circular surface coils 3 mm in diameter would yield the optimal relative SNR over the region of interest. Phantom experiments designed to verify modeling results indeed indicated that the small coil could be used to increase relative SNR over a localized area, however, it was determined that the coil necessary to achieve this localization was not practical for in vivo ³¹ P NMR spectroscopy at 2 Tesla. This impracticality arises from the fact that localization comes at the expense of sample volume, and this reduction in sample volume causes a decrease in signal-to-noise.

In order to accomplish the difficult feat of extracting a usable NMR spectrum from the particularly small sample of interest in this experiment, an inductively-coupled doubly-tuned solenoidal resonator was designed, fabricated, tested, and employed. This coil allowed ¹ H shimming and ³¹P acquisition to take place through the same sample coil which is important in in vivo applications such as this one. While the coil exhibits a diminished ¹H channel performance and can not be used for simultaneous multinuclear acquisitions, it is substantially easier to build and tune than traditional doubly-tuned coils and for phosphorus acquisitions, performed as well as similarly constructed single-tuned coils.

Preparation of the physiological sample for this work involved instrumenting a living animal. In the course of this work, two unique preparation techniques were developed including the use of the pericardial space for loading of fluo-3/AM and employment of a unidirectional breathing apparatus for ventilation. The novel method of loading the fluorescent calcium indicator dye fluo-3/AM into the cells of an in vivo, in situ heart was developed by exploiting the unique anatomy of the particular species under study. This technique not only works with fluo-3/AM but should work with the AM ester forms of other fluorescent indicators as well and may be of use in studying the hearts of other species with similar anatomy. The unidirectional ventilation technique developed in this research not only allows for tighter control and maintenance of blood gas levels but also alleviates the need for respiratory gating during an NMR experiment. Analysis of arterial blood samples taken during unidirectional and tidal ventilation protocols demonstrated that the former is superior to the latter in maintenance of steady-state blood gas and O₂ saturation as well.

Finally, simultaneous dual optical and NMR spectroscopy were demonstrated on an in vivo turtle model in order to measure intracellular diastolic calcium concentrations and high-energy phosphate metabolic state during hypoxia. This technique was found to increase the amount of information available from a given experiment as well as provide parallel correlated measurements from a single animal experiment. Such parallel acquisition strengthens the conclusions which can be drawn from experiments by eliminating the possibility of subject and/or treatment variability which can arise when separate experiments are performed.

7.2 Conclusions

In this work we have demonstrated, for the first time, simultaneous collection of optical and NMR spectra for the simultaneous monitoring of intracellular [Ca²⁺] and phosphorus energy level in an in vivo preparation. The combination of these two techniques into one experiment allows for correlated measurements to be made which is an improvement from what were previously only inferences drawn from parallel experiments and, often, different preparations. While this work focused on a specific application, it should be noted that the two modalities are extremely compatible and useful for a wide range of data fusion experiments. Furthermore, the versatility of the optical instrument designed in this work make it an excellent candidate for exploring some of these other applications with a minimum of instrumentation changes required. Additionally, the fact that this instrument is capable of collecting optical spectra (and not just single wavelength measurements as has been done by others) makes it a valuable addition to the MR spectroscopy lab. It is expected that the combination of optical and NMR spectroscopic techniques will become increasingly popular as the value of the complementary information available is fully realized.

The special preparatory techniques developed during the course of this work should prove valuable to others studying similar species. We have demonstrated that pericardial bathing is an effective and attractive method for loading the AM ester forms of fluorescent indicators into the cells of the heart which makes possible fluorescent cardiac calcium measurements in an in vivo animal preparation. We have also demonstrated that unidirectional ventilation is an effective means of controlling this animals breathing gas mixture which at once allows a steady-state of blood gases and alleviates the need for respiratory gating during NMR acquisitions thus simplifying the experiment.

Theoretical analysis of the SNR of small surface coils led to the conclusion that an array of appropriately sized small surface coils could be used to effect localization of the NMR signal. However, phantom experiments demonstrated that the coils required for this particular application were sufficiently small that tuning problems at 34.63 MHz and signal magnitude considerations made the use of such an array infeasible at this field strength. Since the diameter of the optimal coil for depth resolution spectroscopy is heavily influenced by the thickness of the tissue slab, a thin tissue slab leads to the requirement of a small coil diameter. The volume visible to such a coil is very small, and hence, so is the number of spins visible to the coil. However, the analysis technique presented applies equally well at higher field strengths and frequencies and could be employed for array design in ³¹P spectroscopy applications at higher field strengths or for ¹ H spectroscopy applications. In either of these cases, signal magnitude per volume is increased leading to an increase in the SNR of a small coil, and tuning issues at higher frequencies are less of a problem. For depth localization with a thicker tissue slab, a larger coil diameter could be used, and this analysis technique could be employed to optimize the size of that coil so long as the assumption that the effective resistance of the coil is copper loss dominated holds.

A final result of this work was that a novel double-tuned inductively coupled coil was designed for in vivo NMR spectroscopy applications. It was demonstrated that this coil was fairly simple to build, easy to tune, sufficient for ¹H shimming, and as good or better than similar single-tuned coils for ³¹ P spectroscopy. Further, the fact that it is inductively coupled makes it a candidate for implantation in chronic in vivo NMR studies. The method employed to design this coil was straight forward and can readily be employed for the design of other similar coils. In fact, the results have already been used to fabricate a larger ¹H/ ³¹P surface coil for use in an in vivo investigation of fetal ovine (sheep) brain hypoxia.

7.3 Future Work

There are a number of areas in which the current research could be extended. These have been broken into four distinct areas under which several sub-tasks have been described.

7.3.1 Dual Spectroscopy Calcium/Phosphorus Investigations

Having demonstrated that the collection of simultaneous ³¹ P NMR and [Ca²⁺] fluorescence data is possible in vivo, the next logical step is to utilize this tool to explore cardiac physiology. One experiment of particular interest would be to utilize this technique to compare the relative degree of anoxia tolerance of Chrysemys picta bellii to that of Trachemys scripta which has previously been noted to be markedly less tolerant to anoxia based on results of similar NMR spectroscopy studies [73]. The added information of calcium regulation could be used to more thoroughly contrast the anoxia response of these two animals.

Another immediate application of this research is to adapt the optical fluorescence system to work with a higher field magnet. High SNR ³¹P NMR spectroscopy on isolated perfused hearts is fairly straight forward when high field strengths are available. Thus adapting the CCD system to this experimental set up would not only increase the amount of data, but could potentially increase the quality of data since at higher field strengths substantially faster NMR acquisitions are practical. Thus a more rigorous time course of the interdependence between phosphorus metabolic state and calcium regulation could be studied.

7.3.2 Surface Coil Arrays for NMR Spectroscopy

Though the surface coil array element arrived at for this work was not practical, such an array might be practical at higher field strengths where ³¹P NMR SNR is higher. In order to investigate the potential for the use of arrays on higher field machines, the first step is to repeat the SNR_rel analysis of Chapter III utilizing appropriate parameters for the higher field system (i.e. operation frequency). Then after the optimal array element size has been determined, fabrication of a small loop coil of this diameter can be used on the high field machine with a small phosphorus phantom to determine if the initial SNR of this coil is high enough to be useful. If this is indeed the case, then an array of such coils should be able to be used for spectral localization in phosphorus NMR experiments.

7.3.3 Investigation of Animal Preparation Techniques

Though the special animal preparation techniques developed in this work were briefly characterized in Chapter VI, a more rigorous exploration is probably warranted if they are to be employed regularly. Specifically of concern is the extent to which the pericardial loading technique allows AM-ester dyes to penetrate the heart. It may indeed be the case that only the first few outermost cell layers are being loaded with the dye and that the fluorescence signal detected is representative only of these cells and not of the myocardium in general. A fairly simple experiment is to perform pericardial loading and then excise the heart for investigation under a fluorescence microscope. Transverse sectioning of the myocardium should allow for quantification of the depth profile of fluorescence intensity and hence degree of penetration of the dye. Also of interest is whether the cells at the surface of the myocardium are indeed representative of the myocardium in general, in which case localization of the dye is not a concern.

The unidirectional breathing technique employed here was convenient in that it allowed for continuous ventilation without respiratory gating of NMR acquisitions. Though blood gas experiments tended to verify that the technique was effective, a definite sensitivity to the flow rate and pressure was noticed. In fact, with high flow rates, the lungs of the animal may have become significantly expanded that pulmonary blood flow was restricted resulting in hypoxia even when oxygen-rich breathing gas mixtures were used. Much of the difficulty in using this technique stemmed from the crude mechanism by which flow and pressure were controlled. A precision gas regulator with the ability to operate at very small pressures and flow rates would greatly simplify use of this technique. Further, the availability of such a device would allow for a series of experiments to be conducted in which a maximally effective ventilation pressure could be established. Finally, analysis of blood samples on a blood-gas analyzer more suitable for reptilian blood would strengthen the results of these measures.

7.3.4 Double-Tuned Coil

The double-tuned coil used in this work was extremely effective for in vivo spectroscopy. In fact, the design simplicity of this coil has led to its use in another animal experiment involving ³¹ P NMR spectroscopy of the brains of fetal sheep. An interesting feature of this coil is the fact that it is inductively coupled and as such there are no physical connections between the coil feeds and the sensing coil. It is therefore possible that the sensing coil and the small resonating circuit could be surgically implanted entirely within the animal and external pick-up coils used to communicate RF signals. Construction of an implantable coil would require that the tuning electronics associated with the sensing coil be miniaturized and ruggedized to survive implantation. Suitable pick-up loops would also need to be designed which would still be effectively coupled to the sesnsing coil even with the increased separation. In the case of the turtle, a small trephine hole drilled in the plastron would allow placement of the sense coil and resonating electronics. This hole could then be sealed with a similarly sized plastic cap. Chronic NMR measurements from the heart of the animal would then be possible by attachment of external pick-up coils.