Contact

Dr. Nikolai Avdievitch

Address: Spemannstr. 41
72076 Tübingen
Room number: 1.B.03
Phone: +49 7071 601 713
Fax: +49 7071 601 702
E-Mail: Nikolai.Avdievitch

 

Picture of Avdievitch, Nikolai, Dr.

Nikolai Avdievitch

Position: Senior Research Scientist  Unit: Henning Scheffler

Optimization of Transmit and Receive Performance of a Transceiver Phased Array

Motivation

In spite of great benefits offered by ultra-high field (UHF, > 7T) MRI, its further developments toward clinical applications are associated with significant technical issues, which include low transmit (Tx) efficiency and strong inhomogeneity of the RF magnetic field, B1+ (1). To improve Tx-efficiency and provide for a full coverage of the human brain, UHF multi-element multi-row arrays (2-6) have to be used in combination with 3D RF shimming or parallel transmission (pTx). There are two major types of human head array coils currently developed for the full-brain imaging at UHF. First type, a so called transmit-only/ receive-only (ToRo) array, consists of two nested layers of elements (5). Inner layer, which is positioned closer to a head, contains multiple elements (commonly loops) used only during reception, i.e. receive (Rx)-only elements. Larger array, positioned in the outer layer, provides for local transmission. Second design, a so called transceiver (TxRx) array, consists of a single layer of elements used during both transmission and reception (2-4,6). The Rx-performance of a human head coil can be easier optimized within the frame of the ToRo-design. Current state of the art UHF ToRo-arrays contain 30 and more Rx-only elements (5), which is critical for high SNR and parallel Rx-performance. In turn, TxRx-arrays provide for more efficient transmission, B1+/√P, due to tighter fit and, thus, higher tissue loading. TxRx-design is also simpler then the ToRo-design since it does not require detuning of elements. However, a drawback of the TxRx-design is that the number of elements in an array is limited by the number of available high power RF Tx-channels (commonly 8 or 16). My work focuses on development of new UHF arrays, which combine benefits of both designs, i.e. provide for efficient transmission and the same time do not compromise SNR and parallel receive performance.

Transmit efficiency

Previously we demonstrated that at 400 MHz a 16-loop double-row (2x8) tight-fit human head TxRx-array can be well decoupled entirely by overlapping surface loops without any additional decoupling strategy (7). This based on our analytical findings that at 400 MHz the mutual resistance between adjacent overlapped loops combined within this specific array geometry is minimal, and both the mutual inductance and mutual resistance can be minimized at the same time (8). As a result, the constructed 2x8 array demonstrated very good decoupling, a full brain coverage, and Tx-efficiency, which was 50% higher than that of the larger ToRo-array of similar length (7). Adding two perpendicular TxRx-loops at the superior position of a head (Fig.1A) improves uniformity of the array transmit B1+ profile (Fig.2).

Receive Performance

In the next step we developed a method of increasing the number of Rx-elements in a TxRx-array without compromising the array’s Tx-performance (9). First, we constructed a human head array prototype, which consisted of a single row of 8 TxRx surface loops circumscribing a head, and 8 Rx-only “vertical” loops positioned in the center of each TxRx-loop such to produce B1+ field perpendicularly to magnetic field B0 (similarly as in Fig.1B). All vertical loops were actively detuned during transmission and preamplifier decoupled during reception. Evaluation of the new array demonstrated that neither Tx-efficiency nor maximum SAR was compromised due to addition of Rx-only vertical loops (9). At the same time SNR was substantially improved (~30% near the brain center). Following prototyping 16-loop array, we extended this idea by constructing 32-element tight-fit human head array consisted of 18 TxRx-loops (Fig.1A) and 14 Rx-only vertical loops (Fig.1B) with the total number of Rx elements equal to 32. The general idea of this approach is that the total number of array elements should not exceed the number of available Rx-channels, i.e. 32 in our case. During designing, first, the required number (e.g. 18, Fig.1A) of TxRx-loops is placed around the object tightly to provide for high Tx-performance. Then, the rest of the loops are used as Rx-only elements, which are positioned to minimize interaction with the TxRx-loops, e.g. “vertically” (Fig.1B). In comparison to the common ToRo-design, this method preserves tight fit of the TxRx-loops and, thus, does not compromise the Tx-performance. In addition, it minimizes the total number of array elements as well as the number of active detuning circuits, which are not required for the TxRx-elements. As seen from Fig.3, the ToRo-array still provides for somewhat higher peripheral SNR.

At the same time, the new 32-channel tight-fit array has substantially (~30%) higher SNR near the center of the brain. It is well known that increasing the number of surface loops in a helmet human-head Rx-array only improves peripheral SNR, while SNR near the center practically doesn’t change (10,11). Combination of surface and vertical loops improves SNR near the center. Overall, the new tight-fit array demonstrates high SNR over the entire brain and a substantially more uniform SNR distribution than that of the ToRo-array. Both arrays provide for comparable Rx parallel performance.

References: 1) Vaughan JT et al, Magn Reson Med, 46:24-30, 2001. 2) Adriany G et al. Magn Reson Med 2010; 63(6):1478-1485. 3) Avdievich NI et al, Appl Magn Reson 2011, 41(2):483-506. 4) Gilbert KM et al, Magn Reson Med 2012;67:1487–1496. 5) Shajan G et al, Magn Reson Med, 71:870-879, 2014. 6) Avdievich NI et al, NMR in Biomedicine. 2017;30(2):1-12. 7) Avdievich et al, Proc. ISMRM 25, 2017, 759. 8) Avdievich et al, NMR in BioMed 2017, DOI: 10.1002/nbm.3759. 9) Avdievich et al, Proc. ISMRM 25, 2017, 4309. 10) Wiggins GC et al, Magn Reson Med 2009;62:754-762. 11) Vaidya MV et al, Conc Magn Reson Part B 2014; 44B(3):53–65.

Optimization of Transmit and Receive Performance of a Transceiver Phased Array

Motivation

In spite of great benefits offered by ultra-high field (UHF, > 7T) MRI, its further developments toward clinical applications are associated with significant technical issues, which include low transmit (Tx) efficiency and strong inhomogeneity of the RF magnetic field, B1+ (1). To improve Tx-efficiency and provide for a full coverage of the human brain, UHF multi-element multi-row arrays (2-6) have to be used in combination with 3D RF shimming or parallel transmission (pTx). There are two major types of human head array coils currently developed for the full-brain imaging at UHF. First type, a so called transmit-only/ receive-only (ToRo) array, consists of two nested layers of elements (5). Inner layer, which is positioned closer to a head, contains multiple elements (commonly loops) used only during reception, i.e. receive (Rx)-only elements. Larger array, positioned in the outer layer, provides for local transmission. Second design, a so called transceiver (TxRx) array, consists of a single layer of elements used during both transmission and reception (2-4,6). The Rx-performance of a human head coil can be easier optimized within the frame of the ToRo-design. Current state of the art UHF ToRo-arrays contain 30 and more Rx-only elements (5), which is critical for high SNR and parallel Rx-performance. In turn, TxRx-arrays provide for more efficient transmission, B1+/√P, due to tighter fit and, thus, higher tissue loading. TxRx-design is also simpler then the ToRo-design since it does not require detuning of elements. However, a drawback of the TxRx-design is that the number of elements in an array is limited by the number of available high power RF Tx-channels (commonly 8 or 16). My work focuses on development of new UHF arrays, which combine benefits of both designs, i.e. provide for efficient transmission and the same time do not compromise SNR and parallel receive performance.

Transmit efficiency

Previously we demonstrated that at 400 MHz a 16-loop double-row (2x8) tight-fit human head TxRx-array can be well decoupled entirely by overlapping surface loops without any additional decoupling strategy (7). This based on our analytical findings that at 400 MHz the mutual resistance between adjacent overlapped loops combined within this specific array geometry is minimal, and both the mutual inductance and mutual resistance can be minimized at the same time (8). As a result, the constructed 2x8 array demonstrated very good decoupling, a full brain coverage, and Tx-efficiency, which was 50% higher than that of the larger ToRo-array of similar length (7). Adding two perpendicular TxRx-loops at the superior position of a head (Fig.1A) improves uniformity of the array transmit B1+ profile (Fig.2).

Receive Performance

In the next step we developed a method of increasing the number of Rx-elements in a TxRx-array without compromising the array’s Tx-performance (9). First, we constructed a human head array prototype, which consisted of a single row of 8 TxRx surface loops circumscribing a head, and 8 Rx-only “vertical” loops positioned in the center of each TxRx-loop such to produce B1+ field perpendicularly to magnetic field B0 (similarly as in Fig.1B). All vertical loops were actively detuned during transmission and preamplifier decoupled during reception. Evaluation of the new array demonstrated that neither Tx-efficiency nor maximum SAR was compromised due to addition of Rx-only vertical loops (9). At the same time SNR was substantially improved (~30% near the brain center). Following prototyping 16-loop array, we extended this idea by constructing 32-element tight-fit human head array consisted of 18 TxRx-loops (Fig.1A) and 14 Rx-only vertical loops (Fig.1B) with the total number of Rx elements equal to 32. The general idea of this approach is that the total number of array elements should not exceed the number of available Rx-channels, i.e. 32 in our case. During designing, first, the required number (e.g. 18, Fig.1A) of TxRx-loops is placed around the object tightly to provide for high Tx-performance. Then, the rest of the loops are used as Rx-only elements, which are positioned to minimize interaction with the TxRx-loops, e.g. “vertically” (Fig.1B). In comparison to the common ToRo-design, this method preserves tight fit of the TxRx-loops and, thus, does not compromise the Tx-performance. In addition, it minimizes the total number of array elements as well as the number of active detuning circuits, which are not required for the TxRx-elements. As seen from Fig.3, the ToRo-array still provides for somewhat higher peripheral SNR. At the same time, the new 32-channel tight-fit array has substantially (~30%) higher SNR near the center of the brain. It is well known that increasing the number of surface loops in a helmet human-head Rx-array only improves peripheral SNR, while SNR near the center practically doesn’t change (10,11). Combination of surface and vertical loops improves SNR near the center. Overall, the new tight-fit array demonstrates high SNR over the entire brain and a substantially more uniform SNR distribution than that of the ToRo-array. Both arrays provide for comparable Rx parallel performance.

References: 1) Vaughan JT et al, Magn Reson Med, 46:24-30, 2001. 2) Adriany G et al. Magn Reson Med 2010; 63(6):1478-1485. 3) Avdievich NI et al, Appl Magn Reson 2011, 41(2):483-506. 4) Gilbert KM et al, Magn Reson Med 2012;67:1487–1496. 5) Shajan G et al, Magn Reson Med, 71:870-879, 2014. 6) Avdievich NI et al, NMR in Biomedicine. 2017;30(2):1-12. 7) Avdievich et al, Proc. ISMRM 25, 2017, 759. 8) Avdievich et al, NMR in BioMed 2017, DOI: 10.1002/nbm.3759. 9) Avdievich et al, Proc. ISMRM 25, 2017, 4309. 10) Wiggins GC et al, Magn Reson Med 2009;62:754-762. 11) Vaidya MV et al, Conc Magn Reson Part B 2014; 44B(3):53–65.

WORK EXPERIENCE:

Feb 2017-Current

Senior Research Scientist (100%), Institute of Physics, Ernst-Moritz-Arndt University Greifswald, Greifswald, Germany.

 

May. 2017-Current

Senior Research Scientist (20%), Department for Ultra-High Field MRI (Department chair Prof. K. Scheffler), Max Planck Institute for Biological Cybernetics, Tübingen, Germany.

 

Jan. 2013- Feb. 2017

Senior Research Scientist, Department for Ultra-High Field MRI (Department chair Prof. K. Scheffler), Max Planck Institute for Biological Cybernetics, Tübingen, Germany.

 

July 2012- Dec 2012

Senior RF Engineer, Resonance Research Instruments, Inc., Billerica, MA, USA.

 

Jan. 2007-June 2012

Associate Research Scientist, Magnetic Resonance Research Center, Department of Neurosergery, Yale University, New Haven, CT, USA.

 

Oct. 2000-Dec. 2006

Associate in Radiology, Magnetic Resonance Research Center, Department of Radiology, Albert Einstein Medical College, Bronx, NY, USA.

 

Feb. 1998-Oct. 2000

Electron Spin Resonance (ESR) Lab Manager, ESR Research Resource Center, Department of Physiology & Biophysics, Albert Einstein Medical College, Bronx, NY, USA.

 

1994-Feb. 1998

Postdoctoral Associate, Department of Chemistry, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA.

 

1993-1994

Researcher, Laboratory of Spin and Magnetic Phenomena, International Tomography Center, Russian Academy of Science (Siberian Branch), Novosibirsk, Russia.

 

1992-1993

Researcher, Laboratory of Magnetic Phenomena, Institute of Chemical Kinetics and Combustion, Russian Academy of Science (Siberian Branch), Novosibirsk, Russia.

 

1987-1992

Junior Researcher, Laboratory of Magnetic Phenomena, Institute of Chemical Kinetics and  Combustion, Russian Academy of Science (Siberian Branch), Novosibirsk, Russia.

 

1985-1987

Engineer, Laboratory of Physical Methods of Chemical Kinetics, Institute of Chemical Kinetics and Combustion, Russian Academy of Science (Siberian Branch), Novosibirsk, Russia.

 

1983-1985

Research Technician, Laboratory of Physical Methods of Chemical Kinetics, Institute of Chemical Kinetics and Combustion, Russian Academy of Science (Siberian Branch), Novosibirsk, Russia.

 

EDUCATION:

1993

Ph.D. in Physical Chemistry, Institute of Chemical Kinetic and Combustion, Russian Academy of Science (Siberian Branch), Novosibirsk, Russia.

1985

M.S., Chemical Physics, Department of Physics, Novosibirsk State University, Novosibirsk, Russia.

1983

B.S., Radiophysics, Department of Physics, Novosibirsk State University, Novosibirsk, Russia.

Preferences: 
References per page: Year: Medium:

  
Show abstracts

Patent (4):

Avdievich NI, Hetherington HP and Pan JW: Magnetic-Resonance Transceiver-Phased Array that Compensates for Reactive and Resistive Components of Mutual Impedance between Array Elements and Circuit and Method Thereof, US20130271144 A1, (October-17-2013).
Avdievich NI and Hetherington HP: Surface coil arrays for simultaneous reception and transmission with a volume coil and uses thereof, US8030926 B2, (October-4-2011).
Hetherington HP, Pan JW and Avdievich NI: Improved transceiver apparatus, system, and methodology for superior in-vivo imaging of human anatomy, WO2010110881 A1, (September-30-2010).
Avdievich N, Peshkovsky A and Kennan R: Open half volume quadrature transverse electromagnetic coil for high field magnetic resonance imaging, US6980003 B2, (December-27-2005).
Page:  
1, 2, 3, 4

Export as:
BibTeX, XML, Pubman, Edoc, RTF
Last updated: Monday, 22.05.2017