|Home | About | Journals | Submit | Contact Us | Français|
A four-loop segmented surface coil resonator (SSCR) with electronic frequency and coupling adjustments was constructed with 18 mm aperture and loading capability suitable for in vivo Electron Paramagnetic Resonance (EPR) spectroscopy and imaging applications at L-band. Increased sample volume and loading capability were achieved by employing a multi-loop three-dimensional surface coil structure. Symmetrical design of the resonator with coupling to each loop resulted in high homogeneity of RF magnetic field. Parallel loops were coupled to the feeder cable via balancing circuitry containing varactor diodes for electronic coupling and tuning over a wide range of loading conditions. Manually adjusted high Q trimmer capacitors were used for initial tuning with subsequent tuning electronically controlled using varactor diodes. This design provides transparency and homogeneity of magnetic field modulation in the sample volume, while matching components are shielded to minimize interference with modulation and ambient RF fields. It can accommodate lossy samples up to 90% of its aperture with high homogeneity of RF and modulation magnetic fields and can function as a surface loop or a slice volume resonator. Along with an outer coaxial NMR surface coil, the SSCR enabled EPR/NMR co-imaging of paramagnetic probes in living rats to a depth of 20 mm.
Increasing interest in application of EPR spectroscopy and imaging techniques for the study of the kinetics and spatial distribution of free radicals and oxygen in in vivo biological tissues and animals has generated a great demand for development of EPR sample resonators capable of accommodating large lossy biological samples [1–8]. The need to achieve the highest sensitivity in the presence of motional noise from breathing and animal movements necessitated the development of resonators with active compensation of frequency and coupling by employing circuitry for Automatic Tuning Control (ATC) and Automatic Coupling Control (ACC). Surface coil resonators (SCR) with ATC and ACC utilizing electric coupling circuit designs have been extensively researched and developed [9–11]. These can be used with the sample placed enface as surface resonators or with sample placed within the loop as a slice selective volume resonator. These employ electric coupling circuits of symmetrical design allowing the widest range of frequency and coupling adjustment by means of voltage adjustment applied to corresponding varactor diodes. In practical applications of such resonators for in vivo EPR spectroscopy and imaging, high-level compensation of motional noise was achieved [9–11]. However, this design has certain limitations. In order to perform EPR measurements on larger animals and humans, further development and improvements in surface coil or volume coil resonator design are required to address several critical needs. These include:
Constructing the surface loop resonator with increased aperture raises several problems which must be overcome including:
The four-loop segmented surface coil resonator (SSCR) and coupling circuit described below was developed to address these concerns and circumvent these limitations.
To overcome the inherent size limitations of the single loop coaxial cable resonator, a novel multi loop coaxial cable surface resonator was developed and tested. The general concept of this resonator is schematically shown in Fig. 1 (a four-segment resonator is shown). The inductive loop of this resonator is formed by four identical loop segments (a) connected in series with respect to RF current in the loop. This allows increasing the diameter of the resulting loop. In principal, the number of loop segments may be increased to attain larger diameter at a given frequency. This number must be even for symmetry of the resonator circuit with respect to the signal ground. An additional benefit of this design is its greater mechanical stability than a conventional single loop resonator whose rigidity is based only on a single pair of semi-rigid coaxial cables . Implementation of the three-dimensional SSCR design results in enhanced rigidity and reduced susceptibility to acoustical/vibrational interference from living animals or external sources.
The identical loops formed by conductors “a” and “b” are connected in parallel with respect to the coupling circuit. Fig. 2 shows the schematics of the resonator with coupling circuit. Four segments of the loop are comprised of four identical elements shaped from 50-Ω semi-rigid coaxial cable. (Micro-Coax - UT-141C PTFE non-magnetic cable with copper jacket and center conductor; outer diameter 3.6 mm, center conductor diameter 0.9 mm.) For each element the outer coaxial shield was removed in the center region at 23 mm length (element “ a” at Fig. 1A). The original Teflon dielectric around the unshielded sections was left intact in order to prevent a contact of the wire with the sample and avoid potential contamination. The equal lengths of the cables on both sides of the exposed center conductor are 110 mm each. The lengths are such that each element forms a 3/4 wavelength resonance circuit with serially connected capacitors at the resonance frequency of the resonator, approximately 1.1 GHz. Final coaxial sections length adjustments were performed to achieve the exact desired frequency. Combined together, the four elements resonate at this frequency and form a loop of ~20 mm in diameter with usable diameter of 18 mm. The resulting structure was integrated within a glass epoxy composite body (element “c” in Fig. 1). The thin silver conducting coating was applied to the body to provide RF shielding while being transparent to the modulation frequency magnetic field. The hardened epoxy resin has high dielectric loss coefficient, and silver metallization was used to prevent RF electric field penetration into its volume and degradation of the resonator Q-factor due to dielectric losses in the epoxy volume.
By carefully designing the resonator structure in a way that no short circuits with respect to the modulation field exist, we have achieved “transparency” and homogeneity of the modulation field. The plane of the surface loop is perpendicular to the vector of the modulation field, and this minimizes modulation field distortion and coupling interferences with individual loops of the resonator. To further decrease these interferences each individual loop is connected to the balun through 1 pF capacitors (see Fig. 2). This arrangement serves as a high-pass filter and suppresses penetration of the field modulation frequency signal into the RF signal path.
In addition, by preventing the induction of eddy currents, the problems stemming from eddy current heating and vibrations caused by electro-mechanical coupling are largely suppressed.
This multi-loop coaxial cable resonator is conceptually analogous to multi-gap loop-gap resonator designs. Field homogeneity and EPR signal to noise ratio analysis has been analyzed by Piasecki et al. for multi-gap loop-gap resonators . It was shown that a proportional gain in signal intensity and homogeneity can be obtained as a function of number of gaps for lossy samples.
The coupling circuit used with this resonator uses electric coupling with varactor tuning and coupling provisions similar to the design reported by Hirata et al. [8,9]. Non-magnetic Silicon varactors (Infineon BB833) with 0.75–9.3 pF range of capacitance adjustment over 1–28 V control voltage range were employed. Parallel multi-feeder coupling allowed significant increase in loading capability necessary for achieving near-critical coupling to large lossy samples. In addition, the high symmetry of this coupling arrangement improves the homogeneity of the B1 distribution of the resonator loaded with the lossy sample. In the present resonator, this design is modified to match four loops with the insertion of eight 1 pF capacitors as shown in Fig. 2.
The equation for the resonator frequency is as follows: Where
This allows us to estimate the physical parameters of the resonator for a given frequency and number of loops.
The described segmented surface coil resonator design utilized four identical series connected segments. The capacitance of the individual gap is determined by the length of the shielded part of the coaxial cable. In our case, a 50 Ω cable has 1 pF capacitance per 10 mm length at 1.1 GHz frequency. So, to obtain 0.96 pF capacitance we took 9.6 mm length. This resulted in 0.48 pF capacitance for each gap as two cables connected in series. The addition of a coaxial cable feeding line equal to one half wavelength for phase inversion did not change the resonance frequency.
Equivalent inductance is determined by the geometry of the gap area, which was approximated as square consisting of four straight conductors 19 mm long and separated at the corners by 2.12 mm with L = 0.174 µH .
Final coaxial section lengths and capacitor values adjustments were performed to achieve the desired resonance frequency. We did not calculate the Q-factor since it may have a greater margin of error due to multiple factors not considered in the lumped circuit model such as radiation losses.
Sensitivity profile (B1 field distribution) of surface resonators was considered in . RF magnetic field magnitude decreases exponentially as function of the distance from the plane of resonator loop. In the 4.5 mm diameter single loop resonator described in , the practical limit for distance from the plane of resonator surface is reported to be about 3 mm.
We have measured the absolute intensity of the B1 field distribution as a function of distance from the plane of the resonator surface utilizing a small (3.3 mm diameter) brass sphere using perturbing spheres method . The measurement of B1 field distribution was performed in the resonator loaded with 40 mm diameter cylinder filled with 0.5 mm glass spheres. Voids between glass spheres were filled with saline solution. This phantom simulates tissue load and allowed free positioning of the brass sphere. The brass sphere was cemented to a quartz rod and its position was controlled by a micro positioning device.
For sensitivity comparison purposes similar measurements were performed using an 8-mm commercial surface coil. Direct sensitivity comparison with the single loop resonator is not feasible on samples larger or deeper than the loop diameter because single loop resonator can not accommodate or fully cover these larger volumes. The 8 mm commercial surface coil resonator manufactured by Magnetech, GmbH is the closest coil for comparison available to us capable of coupling to the surface of a large lossy sample at this frequency. Frequency shift was monitored using an Agilent 8719ES vector network analyzer at 3.2mW power.
The measurement results are shown in Fig. 3. Compared to the SSCR, the B1 field distribution shows stronger B1 field for the Magnetech resonator at distances smaller than 5 mm from the coil plane. However, at distances more than 5 mm, SSCR demonstrates stronger B1 field and outperforms the single loop gap resonator. From Fig. 3, one can envision that the commercial resonator may have better sensitivity at distance less than 5 mm from the plane of the loop. However, at bigger distances the SSCR will outperform it. It should be noted that the magnitude of B1 field at a point does not serve as an effective criteria of the overall resonator efficiency. Instead, we propose that a better criterion for comparing resonator efficiency of different design and size should include an integral of B1 over the active volume of the resonator. Also, given sample properties will influence the individual sensitivity.
The microwave properties of the resonator were measured using Agilent 8719ES S Parameter Network Analyzer. The Q-factor of the empty critically coupled resonator was 300. The Q-factor of the resonator with an 18-mm diameter cylindrical bottle, 40 mm length, filled with water, inserted into the center of resonator dropped to 120. The same amount of Dulbecco’s phosphate-buffered saline 50% solution sample or 100% normal Dulbecco’s phosphate-buffered sample saline decreased the Q to 55 or 50, correspondingly. It is important that the resonator allowed critical coupling even with lossy samples of considerable volume.
In order to assess the applicability of the resonator for slice selective experiments and plane selectivity, a phantom assembly consisting of a 20-mm long cluster of 86 capillary glass tubes (cut out of 100 µL glass pipettes ID = 1.1 mm, OD = 1.7 mm) bound together into a octagonal shape was constructed. The tubes were filled with 1 mM solution of TEMPONE (2,2,6,6-tetramethyl-4-oxopiperidine-1-oxyl), spin probe commonly used for biological applications [2,18,19]. This phantom allowed testing the sensitivity and homogeneity of B1 field distribution in the plane of the resonator coil with these small discrete sample tubes (Fig. 4).
The usable volume of the SSCR together with homogeneity of the EPR signal intensity over the surface of the four-loop SSCR that is critical for topical applications was evaluated with three-dimensional spatial imaging . To assist evaluation, a special phantom was constructed. The phantom was a 40-mm outer diameter cylinder with flat 0.5 mm thick bottom end surface, providing a minimum 0.5 mm separation between phantom liquid and the resonator. This end surface thickness simulates the distance between resonator and subject tissue typically used in open chest heart measurements . The cylinder was filled with 100 µM solution of TEMPONE (0.06 mT linewidth) in saline. The concentration was chosen in order demonstrate sensitivity in potential in vivo experiments. Image reconstruction was performed as previously described [21,22]. The size of the phantom exceeded the active volume of the resonator and, thus the obtained images represented the active volume of the resonator. Vertical and horizontal slices of a three-dimensional image of this quasi-infinite sample positioned on the surface of the resonator are shown in Fig. 5A. B1 field profiles obtained from images confirm good signal distribution in the XY plane (Fig. 5A) as well as about 10 mm penetration along the Z-direction (Fig. 5B). Usable volume is defined as volume where signal was above noise threshold. Volume with lesser intensity was zeroed.
Comparison of signal distribution between Figs. 4 and and55 demonstrates quite different pattern. For the images shown in Fig. 4, the signal seems to increase with the position moving near the resonator loop. However, from the image shown in Fig. 5A, higher intensities are observed toward the center of the resonator loop. These variations are the result of the effect of the sample structure on the microwave field distribution. This effect is observed when the size of the sample is comparable with the microwave wavelength in the sample. Homogeneous load with high dielectric constant tends to “ concentrate” the magnetic field resulting in higher concentrations observed around the center of the resonator loop. This effect is much less noticeable with the dispersed load of the discrete capillary tubes.
To evaluate the applicability of the resonator for in vivo applications, images of rats injected i.p. with a paramagnetic charcoal oximetry probe suspension were performed. Female rats of about 250 g weight were obtained from Charles River Laboratories (Wilmington, MA). All measurements were performed under ketamine/xylazine anesthesia (200 mg/kg and 4 mg/kg, respectively, i.p.). The anesthesia was maintained for a period of ~1 h – enough to acquire a series of EPRI and proton MRI images. The rats were prepared by injecting of 1.5 cc of suspension of the charcoal EPR probe (Activated Carbon, Darco® KB-B, −100) in 1:10 proportion by weight with aqueous solution of 5% glucose. Imaging was performed on the EPR/NMR co-imaging system with modified resonator setup . The SSCR was positioned coaxially in the plane of the MRI coil to exclude the need for the moving stage previously required for positioning separate EPRI/MRI resonators. The rat was placed on its side and the resonator was placed over the abdomen in the region of the injection. This arrangement allowed efficient coupling with both EPR and NMR resonators on the living rat. Resultant EPR/MRI images demonstrate that the SSCR allows mapping of paramagnetic density up to 20 mm in tissue depth (Fig. 6).
A four-loop segmented surface coil resonator (SSCR) design with loop diameter of 20 mm and increased aperture, for EPR spectroscopy and imaging of biological samples at 1.1 GHz was designed and constructed. This resonator has greater depth of penetration than conventional single loop resonators, for this frequency and it allows detection from deeper layers up to 20 mm in tissue depth. The resonator has good RF magnetic field homogeneity that was achieved by symmetrical multi-loop geometry. Both mechanical and electrical designs of the resonator provide high transparency and uniformity for magnetic modulation field, as well as mechanical rigidity. The design provides high sensitivity, ease of sample access, excellent electrical stability, and good B1 field homogeneity suitable for in vivo EPR spectroscopy and imaging. Effective separation of the electric component of the microwave field from the active volume decreases Q-factor degradation in the presence of a lossy biological sample, hence, contributing to better sensitivity and coupling capability. Resonator ATC and ACC circuitry containing varactor diodes provided electronic coupling and tuning over a wide range of sample loading conditions. The SSCR design can be used for surface applications or as a slice selective volume resonator and enables larger objects to be studied than possible with conventional SCRs. The open structure and enhanced distribution of RF field homogeneity of the SSCR over a surface or slice volume makes it suitable for EPR spectroscopy and imaging of in vivo animals and isolated organs at higher microwave frequencies.
This work was supported by NIH Research Grants EB0890 and EB4900.