Comparing Saddle, Slotted-tube and Parallel-plate Coils for Magnetic Resonance Imaging

MEASUREMENT SCIENCE REVIEW, Volume 14, No. 3, 2014 171 Compa ing Saddle, Slo ed- ube and Pa allel-pla e Coils o Magne ic Resonance Imaging D. Nespo 1 , K. Ba usek 2 , Z. Dokoupil 2 1 Depa men o Theo e ical and Expe imen al Elec ical Enginee ing, B no Uni e si y o Technology, Kolejni 2906/4, 612 00 B no, Czech Republic 2 Ins i u e o Scien i ic Ins umen s, Academy o Sciences o he Czech Republic, K alo opolska 147, 612 64 B no, Czech Republic The pape is conce ned wi h a compa ison o he p ope ies o RF coils o h ee con igu a ions o MRI measu emen s on small animals. In compa ison wi h he classical saddle coil he p oposed modi ica ion o slo ed- ube coil exhibi s iden ical homogenei y o B 1 ield in a la ge space. The pa allel-pla e coil has a sa is ac o y homogenei y o B 1 ield o e he whole in e nal space. The signal- o-noise a io measu ed o all h ee coils is oughly he same and is gi en by he magni ude o RF p e-ampli ie noise. As he slo ed- ube and pa allel-pla e coils ha e a lowe induc ance compa ed wi h he saddle coil, hey can be uned o esonance on he 200 MHz equency e en a la ge dimensions. The esul s show ha he pa allel-pla e coil has e y good p ope ies o he measu emen o small animals. Keywo ds: MRI measu emen s, RF coils, homogenei y o B 1 , nume ical analysis. 1. I NTRODUCTION HE MRI (magne ic esonance imaging) sys em, which can p oduce high-quali y images in an a bi a y c oss sec ion o he human body, has been ecognized as a new powe ul echnique o medical diagnosis and has g adually come o be employed in p ac ical si ua ions. In he MRI sys em, an RF p obe is used o emi a uni o m RF magne ic ield o e he human body and ecei e he magne ic esonance signal om he body o imaging. Se e al kinds o RF p obes ha e been de eloped. The ans e se elec omagne ic bi dcage esona o is commonly used o MRI. The cha ac e is ics o he bi dcage esona o , including he e ec s o a conduc ing shield, ha e been analyzed by using he bounda y elemen me hod. In his wo k a mul iconduc o ansmission line (MTL) model is de eloped and success ully applied o analyze a 12-elemen unloaded and loaded mic os ip line ans e se elec omagne ic (TEM) esona o coil o animal s udies [1]. To p oduce a uni o m B 1 ield along he coil (z-di ec ion), a mul i-sec ion al e na ing impedance mic os ip elemen was in es iga ed in Akgun [2]. The wid hs o he mic os ip line a e a ied o p oduce an al e na ing low-high impedance con igu a ion. Simula ion and expe imen al esul s a e compa ed o mic os ip esonan elemen s. The slo ed- ube esona o (STR) is also used o MRI applica ions [3]-[6]. Many e o s ha e been made o analyze he EM pa ame e s o he slo ed- ube esona o in o de o show he p ope ies o he p obe and design an op imum s uc u e [3], [6]. In Ben Ahmed and Feham [7], [8] igo ous analy ical exp essions o he EM pa ame e s o he shielded STR wi h a ci cula c oss sec ion ha e been ob ained using he ini e elemen me hod (FEM) and me hod o momen s (MoM) in wo dimensions, bu wi hou aking in o accoun he in luence o he hickness o he STR. A saddle RF coil is used in MR sys ems. Single-channel sys ems [9], [10] equi e an a angemen o coils ha co e s he ull ield o iew (FOV) o he imaged objec . I is an impo an conside a ion o ob ain a uni o m and comple e image o he objec . T adi ional clinical 1.5 T magne s a e being eplaced by 3 T, while esea ch magne s o 7 T, 8 T, and 9.4 T e e ed o as high ield magne s a e being used o animal and human expe imen s [7]-[14]. The g ea appeal o high ield MRI is i s imp o ed image quali y and esolu ion [15]. The signal- o-noise a io (SNR) co ela es in app oxima ely linea ashion wi h ield s eng h, being oughly wice as la ge a 3 T as a 1.5 T [16]. A pa allel-pla e wa eguide was buil and used o acqui e images o a heal hy olun ee ’s leg a 3 T on a clinical MR image [17]. Wa eguides ha e been success ully used o gene a e magne ic esonance images a 7 T o whole-body sys ems [18]. F om hese esul s, i has been es ablished ha wa eguides a e only sui able o 7 T sys ems wi h wide bo es o a leas 60 cm. This is mainly due o he cu -o equency o he cylind ical wa eguides used. To o e come his limi a ion, a pa allel-pla e wa eguide was employed since i s cu -o equency depends on he sepa a ion o he pla es. Zanche [19] has in es iga ed he possibili y o inc easing he SNR and homogenei y o he RF magne ic ield B 1 , o cha ac e ize mic o luidic low wi h di ec de ec ion. To enable his, he used a pa allel-pla e RF esona o wi h pu e phase spa ial encoding. A uni o m B 1 and a s ong B 1 pe uni cu en a e c i ical o magne ic esonance imaging. To design a modi ied RF p obe esona o , nume ical simula ion in a wide- equency band is o en equi ed. The pa allel pla e esona o has been well in es iga ed in non-MR esea ch a eas [21], [22], and he magne ic ield B 1 ( ) be ween he pa allel pla es has been shown o be homogeneous [20], [24]. 2. F ORMULATION OF THE PROBLEM The aim o his wo k is o modi y he slo ed- ube esona o [6] and p opose a pa allel-pla e RF coil o MR mic oscopy and measu emen o small animals o plan s. T 10.2478/ms - 2014 - 0023 MEASUREMENT SCIENCE REVIEW, Volume 14, No. 3, 2014 172 Cha ac e is ics such as he homogenei y o he RF magne ic ield and achie able signal- o-noise a io a e be e compa ed wi h he classical saddle coil. Some heo e ical esul s a e compa ed wi h he measu ed da a o con i m he alidi y o he p esen simula ion. To compa e he modi ied slo ed- ube esona o and pa allel-pla e RF coil wi h he classical saddle coil, an analysis o he RF magne ic ield was pe o med o all he abo e coil con igu a ions, using he nume ical analysis. The esul o he analysis was an op imum coil con igu a ion wi h a minimum inhomogenei y o he ield B 1 , which is shown in Fig.1. Fo he pu pose o expe imen al compa ison, h ee MR p obes we e made wi h coils as pe Fig.1. Fig.1. Schema ic o RF coils unde examina ion: a) saddle coil, b) slo ed- ube coil, and c) pa allel-pla e coil. The saddle coil was made o sil e -pla ed coppe conduc o o 1.4 mm in diame e . The slo ed- ube and pa allel-pla e coils we e made o sil e -pla ed coppe oil 0.05 mm hick. The coil angle Φ was op imized ia nume ical calcula ion o ob ain minimum inhomogenei y o B 1 and is Φ = 70°. This alue is in ag eemen wi h he esul gi en in [6]. Fo all he RF coils, he B 1 ield maps we e measu ed using he MRI me hods, and compa ed wi h nume ical calcula ion. Two main c i e ia we e used in he compa ison, namely he magni ude/size o he space o ield inhomogenei y below ∆B 1 < 10 % and 20 %, and he size o he signal- o-noise a io in he image. Fo applica ions o RF coils in MR expe imen s in which phase MR images a e used he maximum phase de ia ion in he image was e alua ed. 3. M ETHOD The nume ical analysis o he EM ield o coils was sol ed in wo ways: i s as an elec omagne os a ic model and hen as an RF model. The concep o magne os a ic nume ical analysis has been desc ibed in [23]. The ad an age o he elec omagne os a ic model is i s simplici y and good solu ion con e gence in he analysis. The elec omagne os a ic model can only be used in cases when he wa eleng h o he EM ield is much la ge han he coil unde analysis. I he magni ude o he EM ield wa eleng h comes close o he coil dimensions, he nume ical analysis esul s will no longe co espond o he ac ual EM ield dis ibu ion. Fig.2. Compa ison o magne ic induc ion dis ibu ion nea ci cula single- u n coil wi h cen e -line leng h 39 mm, calcula ed ia EM and RF nume ical analyses o di e en wa eleng hs λ = 1.5, 0.6 and 0.3 m. Fig.3. Geome y o nume ical RF analysis o plana single- u n coil. Fig.2. gi es he dis ibu ion o magne ic induc ion B o he ci cula single- u n coil o cen e -line leng h 39 mm. The dis ibu ion o magne ic induc ion is shown in a c oss- sec ion pe pendicula o he coil plane. Fig.2. a) shows MEASUREMENT SCIENCE REVIEW, Volume 14, No. 3, 2014 173 magne ic induc ion dis ibu ion sol ed using he magne o- s a ic model. Figs.2. b), c) and d) show magne ic induc ion dis ibu ion sol ed using he RF model o di e en wa eleng hs. I is ob ious om Fig.2. ha o he wa eleng h λ – 1.5 m (200 MHz) he dis ibu ion o magne ic induc ion ma ches up wi h he esul s ob ained using he s a is ical analysis. Fo he wa eleng hs λ = 0.6 m (500 MHz) and λ = 0.3 m (1 GHz), howe e , he dis ibu ion o magne ic induc ance is di e en . The geome y o nume ical RF analysis o plana single- u n coil is gi en in Fig.3. The geome ic con igu a ion o he model is made up o a coil, which is o med by he acuum. The bounda y condi ions o his coil a e de ined as an impedance medium and a e ep esen ed by (1), whe e µ is he pe meabili y, ε is he pe mi i i y, H is he magne ic ield in ensi y, E is he elec ic ield in ensi y, γ is he elec ic conduc i i y, ω is he angula equency, and n is he no mal ec o . 0)( 0 0 =⋅−+× − nEnEHn ω γ εε µµ j (1) EM ield exci a ion is p o ided using a po wi h lumped pa ame e s. The coil is complemen ed wi h addi ional capaci ance o uning he esonance equency o he equency o NMR sys em = 200 MHz. The coil is placed in he compu a ion space and his space ends in pe ec ly ma ched laye s (PML). PMLs p e en e lec ions and hus simula e an open space. In he i s pa o he RF model he Eigen- equency analysis was used o de e mine he i s esonance equency o he coil being modeled. Based on his analysis, he magni ude o addi ional capaci ance was de ined o uning he esonance equency o he equency o he NMR sys em. Addi ional capaci ance was de ined as elemen s wi h de ined ela i e pe mi i i y ε . When sol ing he Eigen- equency analysis, equa ion (2) was used, whe e he imagina y pa ep esen s Eigen- equency, and he eal pa ep esen s a enua ion. δ ω λ + − = j (2) In he second pa o he RF model he wa e equa ion (3), whe e k is he wa e numbe , and he Maxwell equa ion (4) we e used in he analysis o he high- equency elec omagne ic ield o he coils de ined. 0)()( 0 2 0 1 =−−×∇×∇ − EE ωε γ εµ jk (3) 0 = − = × ∇ HE ωµ j (4) The map o ield B 1 was measu ed using he spin-echo echnique wi h he il angle α = 45°. The ield B 1 was calcula ed acco ding o he ela ion γτ )a csin( 0 1 S S B= (5) Whe e S and S 0 a e he in ensi ies in he image o il angles o 45° and 90°, espec i ely. 4. E XPERIMENT The esul o calcula ing he map o magne ic induc ion B 1 (x, y) o all he coils acco ding o Fig.1. and using he EM me hod is gi en in Fig.4. The maps o B 1 (x, y) a e in he cen e o he coil (z = 0) in a plane pe pendicula o he coil axis. Fig.5. gi es he isolines o he ield B 1 , wi h he maps being no malized o he alue B 1 in he cen e o he map. The le els o ield B 1 a e 0.8 B 1 , 0.9 B 1 , B 1 , 1.1 B 1 , and 1.2 B 1 . Fig.4. Map o magne ic induc ion B 1 (x, y) o es coils acco ding o Fig.1. and using he EM me hod: a) saddle coil, b) slo ed- ube coil, and c) pa allel-pla e coil. The expe imen s we e conduc ed on a Magnex 4.7 T 200 mm i.d. ho izon al bo e supe conduc ing magne a 200 MHz a he Ins i u e o Scien i ic Ins umen s. A wa e - cooled Magnex 150 mm i.d. g adien se was employed, d i en by IECO GPA 200-350 ampli ie s, which can p o ide 180 mT/m maximum g adien s. All expe imen s we e pe o med wi h an MR Solu ion console. The da a measu ed we e p ocessed using he MAREVISI and MATLAB p og ams. One sample was measu ed by all he coils, which was a cylinde wi h a diame e o 20 mm and a leng h o 40 mm, illed wi h a solu ion o deionized wa e o he ollowing concen a ion: 1 li e o wa e , 1.2 g ams o NiSO 4 and 2.6 g ams o NaCl in o de o sho en he elaxa ion imes o T 1 = T 2 = 130 ms. The sample akes up he whole space o he RF coil because in o de o compa e he RF coils, we wan o measu e he map o ield B 1 o e he whole c oss sec ion inside he coil. MEASUREMENT SCIENCE REVIEW, Volume 14, No. 3, 2014 174 Fig.5. Simula ion o he B 1 ield dis ibu ion depic ed as equipo en ial le els, a) saddle coil, b) slo ed- ube coil, and c) pa allel-pla e coil. Field le els o B 1 a e 0.8 B 1 , 0.9 B 1 , B 1 , 1.1 B 1 , and 1.2 B 1 . Fig.6. Map o magne ic induc ion B 1 in a) saddle coil, b) slo ed- ube coil, c) pa allel-pla e coil. The spin-echo me hod (T E = 17 ms, T R = 1s) was used o measu e 11 ans e sal slices 2 mm hick and 5 mm apa om each o he . The FOV size/magni ude was 30x30 mm (128 x 128 px). In he calcula ion o he map o B 1 acco ding o (5), wo images wi h il angles α = 45° and 90° we e measu ed. Fig.6. gi es B 1 ield maps o all he con igu a ions o RF coils. Fig.7. shows he isolines o ield B 1 , wi h he ield maps no malized o he alue o B 1 in he cen e o he map. The le els o ield B 1 a e 0.8 B 1 , 0.9 B 1 , B 1 , 1.1 B 1 , and 1.2 B 1 . Fig.7. Equipo en ial le els o B 1 o : a) saddle coils, b) slo ed- ube coils, c) pa allel-pla e coils. E alua ed om he maps o ield B 1 we e he signal- o- noise a io, size o he homogeneous ci cula and ec angula a ea o 10 % inc ease and dec ease o B 1 , and o he maximum phase change in he image. The esul s a e summa ized in Table 1. Table 1. Calcula ed pa ame e s o RF coils Coil Saddle Slo ed- ube Pa allel- pla e Signal o noise a io 57.2 64.4 49.7 Ac i e coil esis ance ( = 200MHz) [Ω] 3.22 0.82 0.51 Induc ance L ( = 200MHz) [µH] 0.23 0.051 0.048 Homogenei y (±10 %): ci cle wi h diame e [mm] 8.9 9.8 20 Homogenei y (±10 %): ec angle dimensions [mm] 17.7 x 4.5 11.5 x 10.4 20 x 20 ∆Φ [ 0 ] 7.4 9.8 3.5 Using he pa allel-pla e coil he spin-densi y MR images o biological issues (chicken wing and eupho bia) aking up he whole space wi hin he coil we e measu ed. The SE me hod was used, wi h echo ime T E = 17 ms, epea ime T R = 3 s, slice hickness 3 mm, ield o iew 40 x 40 mm, 256 x 256 px o he chicken wing and 30 x 30 mm, 128 x 128 px o eupho bia. MEASUREMENT SCIENCE REVIEW, Volume 14, No. 3, 2014 175 Fig.8. Spin-densi y MR imaging o biological issue, a) chicken wing (SE me hod, T E 17 ms, 40 x 40 mm, 256 x 256 px), b) eupho bia (SE me hod, T E = 17 ms, 30 x 30 mm, 128 x 128 px). 4. R ESULTS AND DISCUSSION Signal- o-noise a io is he a io o he mean alue o image in ensi y in a selec ed space o he s anda d de ia ion o noise in he space ou side o he image. I ollows om he measu emen ha he signal- o-noise a io ob ained is oughly he same o all he coils. Table 1 gi es he ac i e esis ances R and induc ances L o all coils wi h inse ed sample, measu ed on an RLC analyze o he equency = 200 MHz. F om he measu emen i ollows ha he image noise is no due o he noise o he ac i e esis ance o RF coil bu is gi en by he p ope ies o he MR p e- ampli ie . The induc ances o slo ed- ube and pa allel-pla e coils a e ma kedly lowe compa ed o he saddle coil. Bo h hese coils can he e o e be uned a he 200 MHz equency e en o la ge dimensions and hey can be used, o example, o measu e small animals. The homogeneous space o he pa allel-pla e coil on a ci cle o 20 mm in diame e is smalle han + 10 % o B 1 . The inhomogeneous phase in a ci cula space o 20 mm in diame e is up o 10°. F om bo h he expe imen al measu emen and nume ical modeling i is e iden ha o he measu emen o small animals and biological objec s he pa allel-pla e coil is o g ea e ad an age han he saddle o slo ed- ube coil. 5. C ONCLUSION The pape p esen s a compa ison o he p ope ies o RF coils o h ee con igu a ions o MRI measu emen s on small animals. The wo newly p oposed con igu a ions, modi ied slo ed- ube and pa allel-pla e coils, ha e unde iden ical homogenei y o ield B 1 a la ge exploi able applica ion a ea ( ield o iew). The wo RF coils ha e he same SNR, which is due o he noise o RF p e-ampli ie . Since he slo ed- ube and pa allel-pla e coils ha e a lowe induc ance compa ed wi h he saddle coil, hey can a la ge dimensions be uned o esonance a he 200 MHz equency e en i hey a e o la ge dimensions. Resul s o he compa ison indica e ha he pa allel-pla e coil is o ad an ageous p ope ies when measu ing small animals. A CKNOWLEDGMENT The wo k desc ibed in he pape was inancially suppo ed by he esea ch p ojec GACR 13-09086S, esea ch plan MSM 0021630513 and p ojec o he BUT G an Agency FEKTS-10-13. R EFERENCES [1] Bogdano , G., Ludwig, R. (2002). A coupled mic os ip line ans e se elec omagne ic esona o model o high- ield magne ic esonance imaging. Magne ic Resonance in Medicine, 47, 579–593. [2] Akgun, C.E., DelaBa e, L., Sohn, S.M., Snyde , C., Ad iany, G., Ugu bil, K., Vaughan, J.T., Gopina h, A. (2009). No el mul i-channel ansmission line coil o high ield magne ic esonance imaging. In IEEE MTT- S In e na ional Mic owa e Symposium Diges , 7-12 June 2009. IEEE, 1425-1428. [3] Benabdallah, N., Benahmed, N., Benyouce , B., Bouhmidi, R., Kheli , M. (2007). Elec omagne ic analysis o he slo ed- ube esona o wi h a ci cula c oss sec ion o MRI applica ions. Physics in Medicine and Biology, 52, 4943–4952. [4] Alde man, D.W., G an , D.M. (1979). An e icien decouple coil design which educes hea ing in conduc i e samples in supe conduc ing spec ome e s. Jou nal o Magne ic Resonance, 34, 425–433. [5] Li, S., Yang, Q.X., Smi h, M.B. (1994). RF coil op imiza ion: E alua ion o B1 ield homogenei y using ield his og ams and ini e elemen calcula ions. Magne ic Resonance Imaging, 12, 1079–1087. [6] Chen, Q., Sawaya, K., Uno, T., Adachi, S., Ochi, H., Yamamo o, E.A. (1994). Th ee dimensional analysis o slo ed ube esona o o MRI. IEEE T ansac ion on Medical Imaging, 13 (4). [7] Ben Ahmed, N., Feham, M. (2006). Analyzing EM pa ame e s o shielded bandline. Mic owa es & RF, 45, 86–92. [8] Ben Ahmed N., Feham, M., Kheli , M., (2006). Analysis and design o a coupled coaxial line TEM esona o o magne ic esonance imaging. Physics in Medicine and Biology, 51, 2093–2099. [9] Hayes, C.E., Edels ein, W.A., Schenck, J.F., Mulle , O.M., Eash, M.J. (1985). An e icien , highly homogeneous adio equency coil o whole-body NMR imaging a 1.5T. Jou nal o Magne ic Resonance, 64, 622–628. [10] Bae lein, B.A., Ozbay, O., Ib ahim, T., Lee, R., Yu, Y., Kanga lu, A., Robi aille, P.M.L. (2000). The heo e ical model o an MRI adio equency esona o . IEEE T ansac ions on Biomedical Enginee ing, 47, 535–546. [11] Vaughan, T., DelaBa e, L., Snyde , C., Tian, J., Akgun, C., Sh i as a a, D., Liu, W., Olson, C., Ad iany, G., S upp, J., Ande sen, P., Gopina h, A., Van de Moo ele, P-F., Ga wood, M., Ugu bil, K. (2006). 9.4T human MRI: P elimina y esul s. Magne ic Resonance in Medicine, 56 (6), 1274-1282. [12] Vaughan, J., DelaBa e, L., Snyde , C., Ad iany, G., Collins, C., Van de Moo ele, P-F., Ri e , J., S upp, J., Ande sen, P., Tian, J., Smi h, M., Ugu bil, K. (2005). RF image op imiza ion a 7T and 9.4T. In In e na ional Socie y o Magne ic Resonance in Medicine, 13, 953. MEASUREMENT SCIENCE REVIEW, Volume 14, No. 3, 2014 176 [13] Ib ahim, T., Lee, R., Robi aille, P. (2001). E ec o RF coil exci a ion on ield inhomogenei y a ul a-high ields: A ield op imized TEM esona o . Magne ic Resonance Imaging, 19, 1339-1347. [14] Smi g, O., Libe da, O., Smekal, Z., Sp lako a-Puko a, A. (2012). A MRI slice segmen a ion and 3D modelling o empo omandibula join measu ed by mic oscopic coil. Measu emen Science Re iew, 12 (3), 74-81. [15] Přibil, J., Gogola, D., De mek, T., F ollo, I. (2012). Design, ealiza ion and expe imen s wi h a new RF head p obe coil o human ocal ac imaging in an NMR de ice. Measu emen Science Re iew, 12 (3), 98-103. [16] Meadowc o , M.D., Zhang, S., Liu, W., Pa k, B.S., Conno , J.R., Collins, C.M., Smi h, M.B., Yang, Q.X. (2007). Di ec magne ic esonance imaging o his ological issue samples a 3.0T. Magne ic Resonance in Medicine, 57 (5), 835–841. [17] Ma zola, P., Oscula i, F., Sba ba i, A. (2003), High ield MRI in p eclinical esea ch. Eu opean Jou nal o Radiology, 48 (2), 165-170. [18] Vazquez, F., Ma in, R., Ma u o, O., Rod iguez, A.O. (2010). Wa eguide magne ic esonance imaging a 3 Tesla. In Join Annual Mee ing ISMRM-ESMRMB. [19] B unne , D.O., De Zanche, N., F oehlich, J., Paska, J., P uessmann, K.P. (2009). T a elling-wa e nuclea magne ic esonance. Na u e, 457 (7232), 994-U2. [20] Zhang, J., Balcom, B.J. (2010). Pa allel-pla e RF esona o imaging o chemical shi esol ed capilla y low. Magne ic Resonance Imaging, 28, 826–833. [21] Bob o , S., McCa hy, M.J. (1999). Va ia ions on he slo ed- ube esona o : Rec angula and ellip ical coils. Magne ic Resonance Imaging, 17 (5), 783–789. [22] Sheen, J. (2008). A dielec ic esona o me hod o measu ing dielec ic p ope ies o low loss ma e ials in he mic owa e egion. Measu emen Science and Technology, 19 (5), 055701. [23] Akay, M.F., P okopenko, Y., Kha ko sky, S. (2004) Resonance cha ac e is ics o whispe ing galle y modes in pa allel-pla es- ype cylind ical dielec ic esona o s. Mic owa e and Op ical Technology Le e s, 40 (2), 96–101. [24] Nespo , D., Ba usek, K., Fiala, P. (2011). Using nume ical analysis o NMR coils op imiza ion. PIERS P oceedings, 12-16 Sep embe 2011. The EM Academy, 72-75. Recei ed Sep embe 26, 2013. Accep ed June 12, 2014.