BACKGROUND) OF THE INVENTION The present invention relates to a magr etic resonance (MR) imaging system, and more particularly, to a radio frequency (RI) coil system for use within an MR imaging system. A magnetic resonance (MR) imaging system provides an image of a patientor other object in an imaging volume based on delected radio frequency (RF) signals fro ,11 precessing nuclear magnetic moments. A mam magnet produces a static magnetic field, or B0 field, over the imaging vc lume. Similarly, gradient coils within the MR imaging system are employed to quickly switch into effect magnetic gradients along mutually orthogonal x,y,z coordinates in the static B0 field during selected portions of an MR imaging data acquisition cycle. Meanwhile, a radio frequencv (Rij coil produces RP magnetic field pulses, referred to as a B1 field, perpendicular to the B0 held, within the imae;ine volume to exckg the nuclei. The nuclei are fherebv excited to precess about an axis at a resonant R-frequency. These nuclear spins produce a spatially-dependent Rf response sigr al when proper readout magnetic field gradients are applied to them. The RF coil also is able to detect the RF response siRTial of the pi'ecessing nuclear spins and forwards the detected signal to the MR miaging systera. Fhe MR imaging system combines the detected RF' response signa s to provide an image of a portion of the body or object in the imaging voume. hi order to produce accurate images., the static Bo magnetic field, the magneti; field gradients and the B, field generated by the RF coil need to be spatially homogeneous over the imaging volume. :Tradit)onally.)to produce homogeneous fields and gradients, the mam magnet and gradient and RF coils have a cylindrical shape which completely surrounds the patient, in such systems, the B0 Field is t>pioally horizontal, runnmg parallel to the iongimdinal axis of the bore of the cylinder. The cylindrical shape and complete eticasement of the patient insure a highly homogeneous imaging volume. The cylindrical configuration is disadvantagetus, however, m that it severely limits access to the patient and the imaging volume. The cylindrical geometry makes it difficult, if not impossible, for a doctor to perform interactive procedures dunng an MR imaging scan. Additionally, many patients find the cylindncal bore of such traditional MR systems to be cramped, restricting the size of patients that can be examined and also causing claustrophobic reactions in seme patients, I'hus, alternatives 13 the traditional cylindrical geometry are needed. In response to this need, open MR imaging systems have been developed, Ir an open MR. system, the imaging volume is; very accessible and open to both a patient and a doctor. This allows the access to the imaging volume for medical procedures, as well as alleviating the claustrophobic reaction of some patients. Some open IV'IR systems utilize two disk-like magnet pole pieces positioned on opposite sides of the imaging volume with a verticalB0field, These systems have gradient coils and Rf coils that are also flat and disk-like in shape. Tliese open MR systems provide a grec.t amount of access for a doctor or patient in the space between the two disk-like magnet pole pieces. Other open MR systems use two torid-shaped magnet pole pieces positioned on opposite sides of the imaging volume. When set up with a horizontalB0ffiagnetic field, the patient'doctor can access the imaging volume through the bore in the toroids or from the side. Smce the magnet pole pieces are toroid-shaped, the corresponding gradient coils and RF coils are required to be similar m shape and flat to iT.aximize the space between pole pieces. Thus, open MR systems alleviate the access and claustrophobia problems inherent with the traditional, closed system, design Open MR systems are disadvantageous, lowever, in that it is more difficult tc produce homogeneous magnetic fields within their imaging volume. In particular, th; required flatness of the Ri- coil and other components in open MR systems. Similarly, because open MR systems do not completely surround a patient, it is difJicult to obtain a high degree of homogeneit in the staticB0magnetic field, the gradient magnetic fields, and the B1 field. One example of a typical open system RF coil is the dual butterfly design, which is especially inhomogeneous close to the conductors. A flat bird cage design in the form of a weel and spoke structure may al;;o be used, but it is also inhomeneou near the conductors. As referred to herein, a system having a B1field) withhigh homogeneity has substantially equal sensitivity to RF signal throughout the imaging volume. WTien there is inhomogeneit) in theB1field, the sensitivity in the inhomogeneous area increases or decreases. This increase or decrease results in more or LJSS RF sigr.ai being detected, resulting in br ght spots or dark spots in the reconstructed ivtR image. So, for example, the irea near the conductors in typical dual butterfly or tlat bird cage designs are more sensitive than the rest of the imaging volmne, thereby creating very bright areas o ho spots in the image. Also, another disadvantage of typical flatRF coil designs is that the sensitivity doe:s not drop off quickly enough outside of the imaging volume, resulting in RF fields outside o the imaging volume affecting the mage. Inside the imaging volume, the B0ield, thegradient fields, and the RF field are designed to be as homogeneous as possile. Outside of the imaging volume, however, the homogeneity is not as conlrolled. As a result, in areas outside of the imging volume, the superposition of inhcmogeneous Bo and B, fields and alinear gradient fields may give rise to aliasing of signal, whers areas outside of the imaging volume generate a signal with the same freq jency as areas inside the imaging volume. These outside signals may be detected and cause bright spots to be generated within the image. By sharply reducing the RF field sensitivity outside of the imaging voliune, nteraction of outside fields uith the fields inside of the imaging volume is reduced. In typical prior art, the path of the return current v.i through the RF shield, in effect straight under the straight conductor. This results m a straight return current path along the center of the coil, where the renun current path is not being used to produce a sharp drop off in sensitivity As such, current fLF coil designs typically result in stray RJF fields outside of the imaging volums, the so-ay fields in combination with the non-linearity of the gradient coil and the inhomogeneity of the magnet can cause signal from far outside the imaging volume to fold into the image. Thus, in MR sy;tems, it is desirable to produce a ver/ sharp drop off in sensitivity outside of the imaging volume so that signal from outsice of the imaging volume does not affect the imags. In addition to these disadvantages, the design of RF coils for open MRI systems have a. number of other constraints. Fcr example, the diameter of the RF coils is typically limited to the diameter of the magnet pole pieces. The diameter of flat RF coils, and their distance to the iso-center of the imaging volume, affects the abilit)' of the c^ils to produce a homogeneous FF field. For example, it is easier to produce a homogeneous RF field inside an imajiing volume when the diameter of the RF coils is equal to or greater than, the distance from the RF coils to the iso-center of the imaging vcltmie. As mentioned above, since the flatoess of the RF coils already pres.ems a homogeneity problem, restricting the allowable diameter of the RF coils adds another degree of difficulty to the problem. Further, flat RF coils are inefficient compared to cylindrically-shaped coils that surround t:ie imaging volume. Since flat RF coils are inefficient, they require a larger power amp than in a comparable closed MR system. A larger power amp is problematic because it adds additional cost to the system. Further, by requiring more power, the specific absorption ratio (SAR) of ths RF fislds generated by the RF coils may increase. As is known to diose of skill in tie art, SAR pertains to the level of electromagnetic energy which can be absorbed by a patient or medical persormel positioned in or close to the transmit RF coil of an MR system. Within the United Statss, for example, SAR limits are set by the Food and Drug Admimstration (FDA). Because of the tight spacing required for open MR systems, as mentioned above, the RF coils must necessarily be fairly close to the patient surface. Thus, the SAR limits can restrict the amount of power permitted to be utilized by the RF coils. SUMMARY OF THE INVENTION The present invention relates to an RF coil (36, 38) for use in magnetic resonance system, comprising a first and second set (82, 94) of a plurality of parallel-connected primary path conductors (72) for passing electrical current to create a B1 field, said first set (82) of conductors positioned substantially parallel to said second set (94) of conductors symmetrically about a centerline (86) dividing said RF coil, characterized by: said B1 field has a desired homogeneity within an imaging volume resulting from said positioning and from a predetermined current amplitude of each conductor within each of said first and second set when said electrical current energizes each of said first and second sets of primary path conductors; and - a first and second return path conductor (104, 106) for transferring said electrical current between said first and second sets (82, 94) of primary path conductors, said first and second return path conductors (104, 106) symmetrically spaced on each side of said centerline (86), said first and second return path conductors having an outward position further away from said centerline than said first and second sets of primary path conductors resulting in a desired sensitivity drop-off of said B1 field outside of said imaging volume when said RF coil is energized by said electrical current. According to the present invention, an RF body coil for use in a magnetic resonance imaging system has a first and second set of primary path conductors connected to a first and second return path conductor for passing electrical current to create a B.sub.l field having a desired homogeneity with an imaging volume. The first set of primary path conductors is spaced parallel to the second set of primary path conductors, symmetrically positioned about a centerline dividing the RF coil. Further, the first and second return path conductors provide a return path for the electrical current and they are positioned radially outward fi-om the first and second set of primary path conductors in order to produce a desired sensitivity drop-off outside of the imaging volume. Each of the conductors within each of the first and second set of primary path conductors is connected in parallel, while both of the first and second return path conductors are connected in series to each of the first and second set of primary path conductors. Preferably, the first and second set of primary path conductors are in a first plane and the first and second return path conductors are in a second plane, where the first plane is closer to the imaging volume and substantially spaced apart from the second plane so as to provide a desired specific absorption ratio (SAR) within the imaging volume. Thus, the desired homogeneity results from the spacing of the conductors and from a predetermined current amplitude ratio between the conductors within the first and second set of primary path conductors. According to another embodiment, the present invention comprises an RF coil system for use with an MR magnet comprising a pair of magnet pole pieces positioned on opposing sides of an imaging volume. The RF coil system includes a pair of RF coil components positioned on opposing sides of the imaging volume, where each of the RF coil components has a plurality of conductive loops. The conductive loops of each of the RF coil components have a first and second set of primary path conductor segments for passing electrical current to create a B.sub.l field having a desired homogeneity within the imaging volume. The first and second set of primary path conductor segments are positioned parallel and symmetric to each other about a centerline in a first plane, where each conductor within each respective first and second set of primary path conductors segments is connected in parallel. The conductive loops also include a first and second return path conductor segment for providing a return path for the electrical current. Each of the return path conductor segments is serially-connected to opposite ends of the first and second set of primary path conductor segments. Each of the first and second return path conductor segments are positioned radially outward firom the primary path conductors in a second plane that is substantially parallel to and spaced apart firom the first plane. The spaced relationship between the planes provides a desired SAR within the imaging volume In each of the conductive loops, each of the first and second set of primary path conductor segments have at least a first straight conductor for passing a first current and a second straight conductor for passing a second current, wherein the first straight conductor is a specified spacing fi-om the second straight conductor. Also, for each conductive loop, the outward radial positioning of the first and second return path conductor segments, a current amplitude ratio of the first current to the second current and the specified spacing between the first and second primary path conductors combine to provide the B.sub.l field with the desired homogeneity within the imaging volume and the desired sensitivity drop-off outside of the imaging volume. BRIEF DESCRIPTION OF THE ACCOMPANYING DRAWINGS FIG. 1 is a schematic diagram representing a magnetic resonance imaging system of the present invention; FIG. 2 is a schematic diagram representing one embodiment a transmit/receive circuit and an RF coil system of the present invention; FIG. 3 is a schematic view of the structure of a posterior RF coil set; FIG. 4 is a schematic view of an I-channel coil of the posterior RF coil set in FIG. 3; Fig. 5 is a top plan view of a first circu.t board containing the primary path conductor segments of the posterior RF coil set of Fig. 3; fig. 6 is a top plan view of a second cijcuit board containing the return path conductor segments of the posterior RF coil set of Fig, 3; Fig. 7 is an exploded tront perspective view of the circuit boards of Figs. 5-6; Fig. S is a cross-sectional view of tne circuit boards of the posterior RF coil set, similar to Fig. 7, within the MR magnet assembly; Fig. 9 is a schematic diagram of an input circuit of the present invention; Fig. 10 is a schematic diagram of an isclation circuit of the present inventior; Fig. 11 is a plot of a projected magneticB1field in the Y-Z plane provided by the embodiment of Fig. 1; Fig, 12 is a plot of a projected magnetic B, field in the X-Y plane provided hy the embodiment of Fig. 1; and Fig. 11 is a plot of the projected magneic field, in dB's, on the Y-axis. DETAILED DESCRIPTION OF THE INVENTION According to the present invention, referring to Fig. 1, a magnetic resonance (MR) imaging system 10 comprises apatiept 1 positioned in an imaging volume l2-in cpen space 14 between a first magnetic pole piece 16 and a second magnetic pole piece 18 of an MR magnet assembly 20. MR magnet assembly 20 comprises a first and second shim disk 22 and 24, respectively adjacent first and second pole piece 16 and 18, to provide a uniform, static magnetic field 26, orB0magnetic field, across imaging volume 12. A gradient amplitler 28 provides power to a first gradient coil set 30 and a second gradient coil set 32, respectively located adjacent shim disks 22 and 24. The energ zed gradient coil sets 30 and 32, each comprising X-, Y-, and Z-axis gradient coils, produce a magnetic field gradient in the specified direction. RF transmitter 34 supplies the necessary power to antenor RF coil set 36 and posterior RF coil set 38, respectively located adjacent gradient coil sets 30 and 32. Energization of anterior and postenor RF coils sets 36 and 38 transmit RF energy to prcduce a circular, polarized B) magnetic iield 40, that rotates perpendicular to the 13r, magnetic fieic 26. TheB1field 40 excites nuclear spins within patient 11 in imaging volume 12. Sandwiched between each RF coil set 36, 38 and each gradient coil set 30, 32 is a first and second RF shield 42 and 4^-, respectively. First and second RF shiijlds 42 and 44 prevent B1 magnetic field 40 from penetrating gradient coil sets 30 anc, 32, diereby containing the RF energy insid; the imaging volume and preventing a los:i of RF energy witliin the gradient coils, Typically, based upon parameters input by an operator 46 through operator console 4S, a general purpose computer 50 activates pulse sequencer 52 to initiate an MB. data acquasition cycle. Pulse sequencer 52 controls the timing and activation of grailient amplifier 28 and RF transmitter 34 that energize RF transmit/receive circuit 53 i:o produce magnetic field gradients and RF energy. The gradient magnetic fields and RF energy excite nuclear spins and cause an MR response signal to be emitted by tissue of patient 11 at a specified image plane vathin imaging volume 12. RF trarsmit/receive circuit 53 receives the emitted MR response signal from imaging volume 12 of patient 11 and forwards the signa. to receiver 54. RF transmit/receive circuit 53 may acquire the emitted MR response signal from anterior and posterior RF coil sets 36 and 38. Altematively, RF transmit/receive circuit 53 may additionally comprise an RF surface coil (not shown) for receiving the MR response signal. Receiver 54 receives and amplifies the emitted MR response signal, and provides this signal to a reconstruction unit 56. Reconstructi on umt 56 produces data for an MR image of patient 11 at the imaging plane. The image data is provided to general pur])ose computer 50 which displays an MR image on operator console 48. An output of operator console 48 may provide the data to a scan converter 58 which changes th? format of the signal and provides it to a display 60. Display 60 may provide the image of the image plane to aid the physician during a medical procedure, such as surgery. Referring to Figs. 2 atid 3, RF coil sets 36 and 38 each comprise multiple coils in a similar quadrature structure that combine to form an RP coil system 61. RF coil set;; 36 and 38 are preferably substantial mirroi images of each other. Being quadrature coils, each RF coil set 36 and 3S coitains an I-channel coil and a Q-channel coil. The currents in the I- and Q-charnel coils for anterior RF coil set 36 are about 180 degrees out of phase, respectively, with the currents in the I- and Q-chanrel coils in posterior RF coil set 38. Similarly, within each RF coil set 36 and 38, the currents in the I- and Q-channel coils are about 90 degrees out of phase with each other. For example, anterior RF coil set 36 coriprises a quadrature structure having 1-charmel coil 62 and Q-channel coil 64 structure lly and electncally positioned at about 0 degrees and 90 degrees, respectively, Similaly, posterior coil set 38 comprises I-channel coil 65 and Q-channel coil 68 structurally and electrically positioned at about 180 degrees and 270 degrees, respectively. The primary path conductors 70 and 72 from I-channel coils 62 and 66 make an angle of about 90 degrees with the primary pati conductors 74 and 76 of the Q-channel coils 64 and 68, As such, the associatec magnetic fields fi'om the I-channel coils 62 and 66 are substantially perpendicular to the associated magnetic fields from the Q-chaniel coils 64 and 68. Thus, these substantially perpendicular magnetic fields result in a subsiantially circular, polarized B, field 40 (Fig- 1) when the coils are dnven with the phase shifting described herein. The structure of each RF coil set 36 and 38, and the relative position of each coil componer t to iso-center 88, have a dramatic affect on the homogeneity of the B field 40 withir imaging volume 12, die sensitivity drop-otT outside of the imaging volume, and the SAR exposure on patient 11. In particular, the spacmg bewveen each of the primary path conductors, combined with their distance to imaging vol ime 12 and the ratio of the current amplitudes m the pnmary path conductors, have a major affect on the homogeneity of the B1field inside the imagmg volume. Also, the posinonmg of the return path conductors and their distance to imaging volume 12 has a drastic effect on the amount of sensitivity drop-off outside the imaging volume. And finally, the distance of the returnm path conductors and the pninary path conductors to the RJF shield and the imaging volume has a significait affect on the sensitivity drop-off, homogeneity within the imaging volume and SAR exposure on patient 11. The structure of each coil 62, 64, 66, 68 is similar. Using posterior I-channel coil 66 as an example, referring to Fig. 4, each coil comprises primary path conductors 72 in which the peak current amplitudes I, and I, can be varied by means of capacitors to provide a homogeneous, circuhr, polarized B, field 40. Preferably, primary path conductors 72 comprise a first set of pnmary path conductor segments 82, electrically connected in parallel. Each coniuctor of the first set of conductor segments 82 is preferably substantially parsllel ro a central axis 86. Central axis lies within a central plane 87 (Fig. 6) that intersects iso-center SS of imaging volume 12. First set of straight conductor segments 82 comprises at least a first primary path conductor segment 90 and preferably a second primary path conductor segment 92. Conresponding to primiaiy path conductor segments 90 and 92, the current amplitude;; I, and I,, and hence the value of capacitors 78-81, may vary depending on the distance of each primary path conductor segment to the imaging volume and the desired hoirogeneity based on the Biot-Savart law, as is explained below. The ratio of the currents I1 and I2 determines the homogeneity within imaging volume 12. As one skilled in the art will realize, however, the amphtude of the current depend;; totally on the required B1 field iimphtude. Therefore, the value of thi; peai. current amplitudes for 1, and I2 will vary with the requirements for each system Primaiy path conductors 72 additionally comprise a second set of pnmary pati concuctor segn-ients 94 disposed symmetric to first set of primary path conductor segments S2 about central axis 86. Second set 94 also comprises first and preferably seco.ad primary path conductor segments 96 and 98, and capacitors 100-103, having similar current :miplitude and capacitive charactteristics as first and second conductor segments 90 and 92 and capacitors 78-81, respectively. The relative positioning of primary path conductor segments 90, 92, 96 and 98 is symmetric about central axis 86. As mentioned above, primary path conductors 72 compnse first and second set of conductor segments 82 and 94. Each set of conductor segments 82 and 94 includc:s at least one conductor segment. To improve the homogeneity of B, field 40 with imaging volume 12, however, each set of condictor segments 82 and 94 preferably includes at least two conductor segments. Increasing the number of conductor segments within each set 82 and 94 improves the ability to achieve a desired level of homogeneity. First and second primary path conductor segments for each set 82 and 94 are respectively positioned a specified distance x1and x2 from central axis 86. The specified distances x1 and x2 from central axis 85 respectively also correspond to a distiince to unaping volume T2. As is discussed below,/the Biot-Savart law uses the values for X1 and X2 in combination with the ratio of currents I, and I2, to determine the desired B, field 40. Accordingly, for each set of primary path conductor segments 82 and 94 in each coil tne distance to central axis 86 and the current amplitude ratio of the conductcrs are set to achieve a desired homogeneity of theB1field with respect to iso-center 88. The homogeneity of the B; field is preferably better than about + 6 dB, more preferably better than about ± 3 dB, e\en more preferably better than about ± 2 dB, and most preferably better than about ± t.5 dB. Similarly, each coil 62, 64, 66, 68 comprises first and second renim path conductors that respectively transfer the cunent between the first and second set of primary path conductor segments. As such, again using posterior I-chaimel coil 66 (Fig. 4) as an example, each of the first and seco id return path conductors 104 and 106 has a current amplitude I3 = I1 + I2 First and second return path conductors 104 and : 06 are in commumcation with primary path conductor segments 82 and 94 through conductive paths 108-1 ll. First and second return path conductors 104 and 106 each return the electrical current from one end of each set of primary path conductor segments 82 and 94 to the opposite :nd of the opposing set of primary path conductors. J;ach of first and second retura pah conductors 104 and 106 is serially-coimected to each set of pnmary path conducrc'r segments 82 and 94 to fonn a continuous ciicuit. In this manner, current enters and exits first and second set of primary path conductor segments 82 and 94 in a similar manner. As a result, the current rlowing in first set of primary path con r() to provide isolation between the channels and rnnimize energy loss. If the I- and Q-channel coils tor each RF coil set 36 and 38 were identical, then the rettim path conductors for each RF coil set would be nght on top of each other. If the return path conductors were superimposed, then a high capacitance between the return path conductors would be created that would possibly give rise to poor isolation, or a noi'.e or energy trant.fer, between the I- and Q-channels. The spacing bervveen the return path conductois, or the difference between rq-is preferably provides an isolation o::' beaer than about -15 dB, and more preferably better than about -20 dB. Thus, first and second return path conductors 104, 116 and 106, 118 for each I- and Q-channel coil, respecrively, are positioned to maximize isolation between channels and to achieve a sharp drop-off in sensitivity outside ofthe imaging volume while achieving the desired homogeneity of the circular, polarised B; field 40 generated by the RF coils. Referring to Figs. 5 and 6, according to he present invention, the primary path conductors are advantageously positioned in a f rst plane 84 that is substantially sepixated from a second plane 114 that contains the return path conductors. The sepcration of the primary and return path conductors beneficially allows the high amplitude current earned by the return path conductors to be moved away from the patient 11, thereby reducing SAR exposure and optimizing the rate of sensitivity dropoff. Meanwhile, the higher amplitude current carrying return path conductors still contribute to tlie homogeneity of B, field 40 and the sensitivity drop-off, Referring to Figs. 5-8, using postenor coils 66 and 68 as an example, preferabiy the pnmary path conductors 72 and '6 are traces on a first circuit board 6 7 (Fig. 7), while the return path conductors 104, 106, 116 and 118 are traces in a secord circuit board 69 (Fig, 8). Ideally, the traces for each channel are placed on the same surface of each circuit board 67, 69. For examp le, I-chaimel primary path conductor; 72 i\ie on the top surface of first circuit board 6', while I-chaimel return conductors 104 and 106 aie on the top surface of second circuit board 69. Similarly, Q-channel conductors 76 and 116, 118, respectively, are on the bottom surface of circuit boards 67 and 69. respectively. As is known in the art, each circuit board 67 and 69 comprises FR4TM material and TEFL0N(1i! material with copper traces etched into the surface. Other structures ma\ be utilized, however, such as strips of copper or other electrically conducting material positioned to match the structure disclcsed herein. The thickness of each circuit board 67 and 69 is relatively thin, prefenhly from about 15 mils to 65 mils (1 mil = 0.001 inch), and thus the conductors on each surface can be considered lo be in substantiaJly the same plane. Refening to Fig. 5, the conductive paths 108-111 from coil 66, for e.xample, include conductive extensions 108a-11la that extend between circuit boards 67 and 69 to coimect the primary- path conductors 72 to the respective return path conductorE 104, 106, Conductive extensions 108a-l 1 la may include rivets and eyelets, pins, wires or other similar devices that can be used t