F

g
n
a
h
Z

MAGNETOM C!
MR

Function Description
System

Tune-up/QA

08106705

© Siemens AG 2005
The reproduction, transmission or use
of this document or its contents is not
permitted without express written
authority. Offenders will be liable for
damages. All rights, including rights
created by patent grant or registration
of a utility model or design, are
reserved.

Print No.: M4-030.850.20.01.02 English

Replaces: M4-030.850.01.02.02 Doc. Gen. Date: 09.06
2 Revision / Disclaimer
1Revision / Disclaimer

Document revision level

The document corresponds to the version/revision level effective at the time of system
delivery. Revisions to hardcopy documentation are not automatically distributed.
Please contact your local Siemens office to order current revision levels.

Disclaimer
The installation and service of equipment described herein is to be performed by qualified
personnel who are employed by Siemens or one of its affiliates or who are otherwise
authorized by Siemens or one of its affiliates to provide such services.
Assemblers and other persons who are not employed by or otherwise directly affiliated
with or authorized by Siemens or one of its affiliates are directed to contact one of the
local offices of Siemens or one of its affiliates before attempting installation or service pro-
cedures.

MAGNETOM C! M4-030.850.20.01.02 Page 2 of 50 Siemens AG

09.06 SMMR Medical Solutions
Table of Contents 3
1- 0Table of Contents

1 _______ Tune-up and QA_________________________________________________ 4

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Tune-up Procedures only with Body-Spine-Array large (since MR 2002B) . . . . . . . . 5
Quality Assurance Procedures with Body Spine Array large . . . . . . . . . . . . . . . . . . . 6
Tune-up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
RF-related Adjustments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Shim Adjustments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Gradient-related Adjustments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Gradient Sensitivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Coil adjustment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Quality Assurance (QA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Coil-dependent QA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
General QA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

2 _______ Changes ______________________________________________________ 47

Changes from V1.0 to V2.0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

Part ‘Prefix’ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Part ‘System’. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Part ‘Software’. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Part ‘Host/Imager(MRIR)’ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Part ‘Control’ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Part ‘RF System’ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Part ‘Power Distribution’ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Part ‘Tune-up and QA’ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

Siemens AG M4-030.850.20.01.02 Page 3 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
4 Tune-up and QA
2-
1Tune-up and QA

This description provides an explanation of the physical background of the individual pro-
cedures performed during the tune-up and the quality assurance measurement.
The more the service software is simplified and automatized, the more important is a good
understanding of such procedures.
The tune-up is a collection of procedures to tune the MR-system after installation.

MAGNETOM C! M4-030.850.20.01.02 Page 4 of 50 Siemens AG

09.06 SMMR Medical Solutions
Tune-up and QA 5

Introduction 0

NOTE The Tune-up and QA procedures are coil-dependent. The general

Tune-up and QA procedures are only available with the Body
Spine Array large.

Tune-up Procedures only with Body-Spine-Array large (since MR 2002B) 0

As with other MAGNETOM system, the tune-up can be divided into the following catego-
ries
• RF-related adjustments
- Tuning Calibration
- TX-Tuning
- RF-Characteristics
• Shim adjustments:
- Mechanical Shim
- Phantom Shim
• Gradient-related adjustments
- Eddy Current Compensation (ECC)
- Gradient Delay
- Gradient Sensitivity
• Coil adjustment
- Image Brightness
It is important not to change the order of the individual steps.

Siemens AG M4-030.850.20.01.02 Page 5 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
6 Tune-up and QA

Fig. 1: Tuneup

Quality Assurance Procedures with Body Spine Array large 0

• Coil Dependent QA (has to be performed with each coil)

- Coil SN Check
• General QA (only with Body-Spine-Array large)
- Receive Path Calibration Check
- Phantom Shim Check
- Image Orientation
- RF-Verify
- Linearity Check
- ECC Check
- Grad Delay Check
- Gradient Sensitivity Check
- Calc Artifacts
- Spike Check
- Stability Check
- Synthesizer Check
- Field Stability

MAGNETOM C! M4-030.850.20.01.02 Page 6 of 50 Siemens AG

09.06 SMMR Medical Solutions
Tune-up and QA 7

Fig. 2: QA

Siemens AG M4-030.850.20.01.02 Page 7 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
8 Tune-up and QA

Tune-up 0

RF-related Adjustments 0

Tuning Calibration
All four preamplifier paths (two coil plugs with two amplifiers each) have to be calibrated
so that the system can measure and adjust the impedance of the connected coil.
This calibration is performed by using the TTX-output of the modulator which couples a
small RF-signal (1 mW) to the input of the selected preamplifier. Depending on the imped-
ance at the input of the preamplifier (open, short or 50 Ω), a certain amount of RF is
reflected, amplified and received at the demodulator.
By means of the three received signals, the system can calibrate itself as a vector imped-
ance meter.

MAGNETOM C! M4-030.850.20.01.02 Page 8 of 50 Siemens AG

09.06 SMMR Medical Solutions
Tune-up and QA 9
Tuning Calibration
Fig. 3:
Siemens AG M4-030.850.20.01.02 Page 9 of 50 MAGNETOM C!
Medical Solutions 09.06 SMMR
10 Tune-up and QA

TX-Tuning
Measurement procedure
The service tuning of the transmit coil is performed by measuring the forward and
reflected signals by means of the directional coupler in the RFPA and the so-called trans-
mission measured at the RX-output of the power splitter in the body tune box (BTB).
These three values are measured using a rectangular RF-pulse with 100 W and 3.5 ms
pulse duration. Furthermore, the measurement is carried out for three frequencies:

flow = 14.60 MHz - 100kHz = 14.50 MHz

fo = 14.60 MHz
fhigh = 14.60 MHz +100kHz = 14.70 MHz

Measurement results
From those measured values, the system calculates an alignment factor AF and a trim-
ming factor TF which are displayed at the display and tune panel (DTP).

Fig. 4: Example of Alignment and Trimming Factor

MAGNETOM C! M4-030.850.20.01.02 Page 10 of 50 Siemens AG

09.06 SMMR Medical Solutions
Tune-up and QA 11

Fig. 5: Example of Alignment and Trimming Factor

Siemens AG M4-030.850.20.01.02 Page 11 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
12 Tune-up and QA
TX Tuning
Fig. 6:
MAGNETOM C! M4-030.850.20.01.02 Page 12 of 50 Siemens AG
09.06 SMMR Medical Solutions
Tune-up and QA 13

RF-Characteristics
General
Slice selection uses selective RF-pulses, where the amplitude of the carrier frequency is
modulated with a kind of sinc-function.
Non-linear components
The envelope of the transmitter pulse is distorted by non-linear components in the trans-
mitter path. If a correction is missing or is insufficient, this will lead to a slice profile deteri-
oration and wrong flip angles, which again deteriorate the S/N-ratio.
Predistortion
Therefore, the RF-characteristic of the whole transmitter path is measured and the trans-
mitter pulse is predistorted with the inverse function of this characteristic. The predistor-
tion of the RF-pulse compensates for the subsequent distortion.
Measurement Sequence
For plotting the curve, a test sequence is used which generates linear ramps. These linear
ramps use the entire dynamic range or the TX-module.
RF-characteristic transfer function
In order to detect the RF-characteristic transfer function, the output of the transmitter path
(high power) is fed back via a directional coupler included in the RFPA into the RX-mod-
ule.

Siemens AG M4-030.850.20.01.02 Page 13 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
14 Tune-up and QA

Fig. 7: RFPA Characteristic

MAGNETOM C! M4-030.850.20.01.02 Page 14 of 50 Siemens AG

09.06 SMMR Medical Solutions
Tune-up and QA 15
RFPA Characteristic
Fig. 8:
Siemens AG M4-030.850.20.01.02 Page 15 of 50 MAGNETOM C!
Medical Solutions 09.06 SMMR
16 Tune-up and QA

Shim Adjustments 0

The specifications for shimming are:

NOTE peak-to-peak < 60 ppm and Vrms < 5 ppm measured in a 360 mm
spherical volume.

The shimming of the MAGNETOM C! system is performed using the following steps:
• Factory shim at SMMR.
• Mechanical shim.
• Passive shim (using permanent magnets) in case the homogeneity is out of specifica-
tion after the mechanical shim. Mechanical shim has to be repeated after passive shim.
• Phantom shim of the linear terms (using gradient offset currents).

Mechanical Shim
The mechanical shim optimizes 4 tesseral terms using bolts at the top and the bottom
rose ring.
Additionally, two plugs in the center of the two pole shoes can be moved up and down to
shim the zonal terms.

MAGNETOM C! M4-030.850.20.01.02 Page 16 of 50 Siemens AG

09.06 SMMR Medical Solutions
Tune-up and QA 17

Fig. 9: Tesseral and Zonal Terms

The mechanical shim consists of 4 steps:

Siemens AG M4-030.850.20.01.02 Page 17 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
18 Tune-up and QA

1. Field Stability Measurement

The field stability is checked measuring 256 FIDs in about 77 s. The field drift during
these 77 s is calculated and extrapolated to 10 min.

Fig. 10: Field stability

MAGNETOM C! M4-030.850.20.01.02 Page 18 of 50 Siemens AG

09.06 SMMR Medical Solutions
Tune-up and QA 19

2. Phantom Shim Measurement

The next step is the measurement of the inhomogeneity by means of a phantom shim
measurement. This results in the ppm-values for the tesseral and zonal terms from
which the corrections are calculated.

Fig. 11: Result of phantom shim

Siemens AG M4-030.850.20.01.02 Page 19 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
20 Tune-up and QA

3. Correction of the Tesseral Terms

The tesseral terms are shimmed by turning 20 threaded bolts at 8 positions.

Fig. 12: Correction of Tesseral Terms

Fig. 13: Position of Terms

MAGNETOM C! M4-030.850.20.01.02 Page 20 of 50 Siemens AG

09.06 SMMR Medical Solutions
Tune-up and QA 21

4. Correction of the Zonal Terms

The Zonal terms are shimmed by moving the top and bottom plugs up or down. A
frequency displayed at the tuning panel tells about the successful adjustment.

Phantom Shim
The Phantom Shim performs a 3D-measurement.
3D-sequences apply an additional phase-encoding table in slice-selection direction. The
total measurement time becomes:

TA=TR*nPE*nSL
With: nPE=number of phase-encoding steps in phase-encoding direction
nSL=number of phase-encoding steps in slice-selection direction

The following graphics show the principle of reconstruction: a third FFT has to be per-
formed to resolve the 3D-volume into single slices.

Fig. 14: 3D volume

Due to measurement time optimization, the matrix is 32x32x32. The sequence is a
3D-DESS-sequence, i.e. two echoes, a FISP and a PSIF echo are generated.

Siemens AG M4-030.850.20.01.02 Page 21 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
22 Tune-up and QA

Fig. 15: 3D-DESS Sequence

The measurement is performed twice, for the second measurement all gradients are
reversed to eliminate eddy current effects.
The echo-time difference of 10.3 ms has the shortcoming that the true absolute off-reso-
nance-frequencies cannot be analyzed: Due to Nyquist Theorem, the highest frequency
that can be analyzed correctly is only ±55 Hz. However, for the check and optimization of
the Bo-field, it is sufficient to compare the field differences between neighboring voxels.
The so-called differential shim equation will try to minimize the difference between:
• the field generated by the three gradient coils
• the measured magnetic field inhomogeneity.
The 8 field terms to be optimized are:

MAGNETOM C! M4-030.850.20.01.02 Page 22 of 50 Siemens AG

09.06 SMMR Medical Solutions
Tune-up and QA 23
Tab. 1 Field Terms (not gradient direction!

Spherical Harmonic Field

A(1,0) z
A(1,1) x
B(1,1) y
A(2,0) z2 - (x2+y2)/2
A(2,1) xz
B(2,1) yz
A(2,2) (x2-y22)/2
B(2,2) xy

Numerically, the problem is a least square fit leading to a system of almost 100,000 linear
equations with 3 unknowns: the 3 gradient offset currents.

Fig. 16: Result of phantom shim

NOTE The gradient offset values are used in the 3D-shim mode.

Siemens AG M4-030.850.20.01.02 Page 23 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
24 Tune-up and QA
Fig. 17: 3D Imaging with 3 Fourier Transformations
MAGNETOM C! M4-030.850.20.01.02 Page 24 of 50 Siemens AG
09.06 SMMR Medical Solutions
Tune-up and QA 25

Gradient-related Adjustments 0

Eddy Current Compensation

Switching gradients produces eddy currents in all conductive surrounding metallic struc-
tures. The eddy fields are opposed to the gradient fields, so that in the beginning of a gra-
dient, the field is too small resp. When the gradient is switched off, it will be followed by a
decaying eddy field. The eddy currents are measured and characterized by their
• time constants
• amplitudes
Spatial Dependency of Eddy Current Fields
Measurements have shown, that for slice shifts that are not too large, it is sufficient to
divide the eddy field into a 0th order and 1st order term.
0th order term
This term arises from an asymmetry of the gradient coil with respect to the pole shoes and
other conductive materials. The field is space independent, hence it is present in the com-
plete imaging volume. It adds up to the normal B0-Field. The amplitude (x) is given in the
unit µTm/mT, i.e. a gradient ramped up to 1mT/m will produce a 0th order term with an
amplitude of x µT. The time constant of the most important part is about 500 ms, but there
are also additional shorter time constant parts.
1st order term
This is the most important term as it has the same symmetry as the gradient field. The
amplitude is given in % of the gradient.
Higher order terms
They are usually small and negligible - as long as the slice shifts and off center zooms are
not too large. No compensation is made for such terms.

Siemens AG M4-030.850.20.01.02 Page 25 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
26 Tune-up and QA

Fig. 18: ECC Terms

Measurements of the Eddy Currents

MAGNETOM C! M4-030.850.20.01.02 Page 26 of 50 Siemens AG

09.06 SMMR Medical Solutions
Tune-up and QA 27

The sequence for the eddy current measurement applies gradients and RF-pulses gener-
ating 23 spin echoes with delay times 0.39 ms to 9000 ms after the gradient.
The small spherical phantom and the body coil are used. The gradient amplitude is 4
mT/m, the spin echoes are generated in 3 mm slices with slice shifts of +50 mm and -50
mm. Hence the gradient amplitude in the slices is
±4 mT/m × 0.05mT = ±0.2mT
which corresponds to a frequency of
±0.2T × 10-3 × 42.577 × 106 Hz/T = ± 8515.4Hz
at slice position +50 mm or -50 mm. This frequency will be displayed as “Reference Fre-
quency” in the measurement files eccx_xxx.dat. An example is given in the following
header:

Fig. 19: ECC File Header

Siemens AG M4-030.850.20.01.02 Page 27 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
28 Tune-up and QA

Fig. 20: ECC Measurement

MAGNETOM C! M4-030.850.20.01.02 Page 28 of 50 Siemens AG

09.06 SMMR Medical Solutions
Tune-up and QA 29

The measurement is made at these moderate slice shifts as we do not want contributions
from higher order terms. After the header, the measurement data follow. Only the first and
the last echoes are given:

Fig. 21: ECC data

The signal Nr.1 is measured for any frequency offset, the signal 2 is measured to correct
for the eddy currents generated by the slice-selection gradient (measurement of long
delays).
The data for a given delay time will be rejected if the amplitude “Ampl“ is less than a given
threshold-value. This happened in signal number 11 in the above table, the frequency
was not calculated for this measurement point. “Points“ in the last column means the
number of ADC-sampling points for the given signal. The shorter the delay time after the
gradient, the shorter the total sampling time and hence the number of sampling points.
The phantom shim has to be made prior to eddy current compensation in order to avoid
signal losses of the echoes. After the sequence was run for the two slice positions +50
mm and -50 mm, the data first will be added up and then subtracted in order to separate
the 0th order term and 1st order term.
The program will then calculate:
• Gradient compensation (for compensation of 1st order term):

Fig. 22: ECC Compensation

Siemens AG M4-030.850.20.01.02 Page 29 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
30 Tune-up and QA

In this example the overshoot tells us that the initial gradient amplitude has to be 1.17 %
higher than the normal value without eddy current compensation.
• Bo - compensation (for compensation of asymmetry)

Fig. 23: ECC B0

The first plot of the uncompensated eddy fields shows a smooth curve. The next iteration
will already establish a large improvement.

Fig. 24: ECC 1st Iteration

The better the compensation, the more small oscillating eddy fields become visible. The
example shows the plot of the 3rd iteration.

Fig. 25: ECC 3rd Iteration

The specification values are given in the table below. Note that the value for delays 2 -
100 ms is a peak-to-peak value which can be much smaller than the values for delays > or
< 10 ms , For image quality not so much absolute value than the change in a given time
interval is important.

MAGNETOM C! M4-030.850.20.01.02 Page 30 of 50 Siemens AG

09.06 SMMR Medical Solutions
Tune-up and QA 31

Fig. 26: ECC result after 3rd Iteration

Compensation of Eddy Current fields
The gradient pulses calculated by the Gradient Wave Form Generator GCTX are evalu-
ated by a special signal processor for eddy current compensation. The gradient waves
are digitally filtered. Four output signals are produced:
• The 3 gradient signals is going to the gradient amplifier and the gradient coil. This pre-
distortion compensation (as usual) for the first order eddy currents.
• A 4th signal,called B0 -signal, goes to the numericallly controlled oscillator, i.e. the syn-
thesizer and updates its frequency every 10µs to compensation for B0 eddy cur-
rent.This compensation is performed for the RF-pulses(modulation) and for processing
of the MR-signals(demodulation).
Eddy Current and Image Quality
Eddy currents insufficiently compensated have numerous effects on the image qual-
ity,which can only be discussed briefly here. The effects of a poorly adjusted gradient
depend on the following:
1. Function of the Gradient: slice selection, phase-encoding or frequency-encod-
ing(read-out)
2. Time constant of the poor adjusted eddy currents in relation to the pulse duration
3. Type of sequence(spin-echo-sequence, gradient-echo-sequence...)
4. Sequence-timing(TR,TE,TI)
Short eddy currents
The effects of incorrect or poorly adjusted short eddy currents are limited to the pulse
duration and to operations directly after the pulses.
Slice selection gradient
For slice selection, RF-energy is already applied, although the slice selection gradient has
not yet reached its full end amplitude. This broadens the slice profiles and also selects
neighboring regions(slice cross talks)
In multi-slice images,this “pre-saturation“ of neighboring slices leads to a reduction of the
S/N-ration, in addition, the S/N -ration will be reduced by a not-yet-complete rephasing of
the magnetization in slice selection direction.
Frequency encoding gradient
The singal is read out although the frequency encoding gradient has not yet reached its
full amplitude. Therefore, frequency encoding starts at the beginning of the readout period
with reduced gradient amplitude.
Each pixel in the slice delivers a lower frequency than it would have with a full gradient
amplitude. This can be observed as blurrings. especially at te edges.

Siemens AG M4-030.850.20.01.02 Page 31 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
32 Tune-up and QA

In a sequence, many gradient pulses are required at certain intervals. compared to the
long time constants of the eddy currents. the gradient pulses lie closely together. Between
the gradient pulses, there is always a “remaining“ gradient field of the long eddy currents
from previous gradient pulses.
Slice selection gradient
A residual gradient field of the slice selecition gradient is still effective during frequency
encoding(=receiving)and,depending on the slice position, causes a frequency offset of the
signal.
In the image, this leads to a shift of the whole object. Since shifting depends on the slice
position. this is especially disturbing in multi-slice imaging with large slice shift.
The residual gradient field between the gradient pulses can be compared to an additional
offset current of the gradient which deteriorates the magnetic field homogeneity in fre-
quency-encoding direction. This slightly distorts the image.
Gradient Asymmetry (B0-term) and Image Quality
Reduction of Signal-to-Noise
The S/N-ration is reduced if the particular gradient with the asymmetry is used as slice
select gradient.

Explanation: Selected slices for the 90° and the 180° pulse are not the
same, since transient fields (asymmetric eddy currents)
change the selected gradient strength.

Image Shift
The image is shifted if the particular gradient with the asymmetry is used as a slice-selec-
tion gradient.

Explanation: For echo and readout times that are small compared to the
typical time constants of the B0-term, the asymmetry causes
a constant additional gradient field during readout.

Gradient Delay
General
Precise timing of gradient pulses and RF is very important for good image quality.
Gradient pulses and analog gradient amplifier, however, lead to a delayed switch-on time
that is slightly different for the three gradient axes. This causes problems, e.g. in paraxial
slices, when at least two gradient axes have to be switched on simultaneously for a cer-
tain function (e.g. slice selection).
Measurement
The measurement uses a spin-echo sequence without phase-encoding, hence only one
line is measured. As usual, the gradient to be tested is switched on once between the 90°
and the 180° pulse and then again during the readout time of the spin echo.
When first switched on, the gradient causes the magnetization to dephase in the direction
of the gradient axis. The amount of dephasing is proportional to the time-amplitude inte-
gral of the gradient pulse. The 180° -pulse inverts the phases, so that the second gradient
pulse after the 180° pulse rephases the magnetization again. The time-amplitude integral
of the read-out gradient prior to the 180° pulse is called GR/2. The maximum echo ampli-

MAGNETOM C! M4-030.850.20.01.02 Page 32 of 50 Siemens AG

09.06 SMMR Medical Solutions
Tune-up and QA 33

tude we get exactly after the same GR/2 integral, now applied by the rephasing gradient
pulse. As we know when this pulse starts, we also can predict the time when the maximal
echo should be received. By determining the real time for the maximal echo amplitude,
the system can calculate the required correction time.

Fig. 27: Gradient Delay Measurement

Siemens AG M4-030.850.20.01.02 Page 33 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
34 Tune-up and QA

Fig. 28: Gradient Delay

The adjustment sequence is repeated with three different gradient amplitudes (1 mT/m,
4.5 mT/m and 8 mT/m). The calculated delays of the three measurements should be less
than 15µs.

Gradient Sensitivity 0

General
The image size of an object depends on the displayed frequency range (selected via FOV
and matrix), which is the result of the gradient amplitude. The gradient amplitude deter-
mines the frequency of the pixels in the object.
The higher the gradient amplitude, the more the resonance frequency of the pixels will
increase. In order to calibrate the value for the sensitivity given in mT/m of LSB (LSB =
least significant bit), an object with well known diameter (large spherical phantom; D =
240 mm) is measured.

MAGNETOM C! M4-030.850.20.01.02 Page 34 of 50 Siemens AG

09.06 SMMR Medical Solutions
Tune-up and QA 35

Fig. 29: Gradient Sensitivity Results

Gradient Amplitude
The gradient amplitude determines the frequency of a pixel in the object. With the gradient
amplitude growing, the resonance frequency of the pixel will increase.
Frequency Range
The frequency range displayed depends on the pixel bandwidth (pbw) and the matrix size
(MA). With this frequency and the selected field of view (FOV), the gradient amplitude can
be determined rather easily.
Example for the Read-out Gradient
The amplitude of the readout gradient (GR) can be calculated in the following way:
∆f = pbw × MARO
∆B = GR × FOV
∆f = γ / 2π × ∆B
pbw × MARO = γ / 2π × GR × FOV
GR = pbw × MARO / (42.577 MHz/T × FOV)
DAC-setting for max. Field
The gradient coils used in the MAGNETOM C! generate a field gradient of at least 20
mT/m at a current of approx. 140 A. This value requires a DAC output of 2.8 V (50 A/V).
For the software, 20 mT/m means a maximum selectable gradient. Consequently, the
GCTX sets all 17 bits of the DAC (1FFFF, MSB is sign bit).

Siemens AG M4-030.850.20.01.02 Page 35 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
36 Tune-up and QA

Fig. 30: Gradient Sensitivity

Correction Value
If the DAC is driven to full output (1FFFF), a gradient current of approx. 140 A will flow,
generating a gradient 20 mT/m. By means of a correction value (gradient sensitivity), the
digital value is adapted to the actual gradient coil sensitivity.
This value should be at least: (20mT/m) / (217-1) = 1.53 × 10-4 mT/m
Center Position of the Phantom
During this measurement the patient table feed is checked as well. This patient table feed
is the distance from the light marker position to the isocenter. In case of the MAGNETOM
C! system, it is in the X-direction.
The phantom has to be centered in all three directions. The allowed tolerance is 10 mm. If
necessary, the patient table and/or phantom has to be repositioned and the measurement
repeated. Afterwards the new value for the patient table offset and the three values for the
gradient sensitivity will be stored in the status file.
Measurement
To determine the gradient sensitivity, three images with transverse, sagittal and coronal
orientation are performed.
Each gradient is switched once as a phase-encoding and once as a read-out gradient.
The average diameter in each direction is put into relation to 240 mm (diameter of phan-
tom) and the corresponding gradient sensitivity value is corrected.

MAGNETOM C! M4-030.850.20.01.02 Page 36 of 50 Siemens AG

09.06 SMMR Medical Solutions
Tune-up and QA 37

Fig. 31: Gradient Sensitivity Results

Coil adjustment 0

Image Brightness
General
This is the only coil-dependent adjustment in the tune-up. All previous adjustments are
so-called general tune-up adjustments which require the Body Spine Array large. With
any other coil only image brightness can be performed.
Measurement

Siemens AG M4-030.850.20.01.02 Page 37 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
38 Tune-up and QA

A spin-echo sequence is used and the brightness in an ROI is evaluated. This size, posi-
tion and brightness specifications are coil dependent.

Fig. 32: Image Brightness

MAGNETOM C! M4-030.850.20.01.02 Page 38 of 50 Siemens AG

09.06 SMMR Medical Solutions
Tune-up and QA 39

Quality Assurance (QA) 0

Coil-dependent QA 0

This QA step has to be performed separately with each coil. The corresponding coil and
phantom setup is shown via photos on the monitor of the MRC after connecting the coil
and starting the QA measurement.

SN DIP
General
Using the SN_DIP_SRM program, the following values will be calculated or checked:
• Signal-to-Noise ratio (S/N);
• Diameter and Intensity Profile (uniformity, DIP);
Signal-to-Noise (SN) Ratio
The signal-to-noise ratio (SNR) is a sensitive but rather unspecific value for the perfor-
mance of the MR-system. Since the sensitivity of all coils is spatially-dependent, the SNR
is measured at a reference point. This reference point approximates the position of the
anatomical feature for which the coil was designed.
Region of interest (ROI)
A square area of the image centered at the reference point defines the region of interest
used to evaluate the signal-to-noise ratio.

Fig. 33: SN_DIP

S/N calculation

1. Standard deviation (SD) between each pixel value and the fit for each line and each col-
umn:
(SDline/column )2 = [1/(n-1)] × ∑(fit-pixel value)2i ( i from 1to n)
2. Calculation of the signal-to-noise (S/N) for each line and each column using the mean
value (of lines / column) over standard deviation:
S/Nline/column = mean valueline/column / SDline/column
3. Final S/N-value is the mean value of all lines and column from number 2
S/NROI = ∑S/Nline/column / ( number of lines + number of column )

Siemens AG M4-030.850.20.01.02 Page 39 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
40 Tune-up and QA

Fig. 34: Signal to Noise Ratio

Fig. 35: SN Ratio Result

Diameter and Intensity Profile (DIP)
Diameter
The diameter of the phantom is measured in the horizontal and vertical direction and com-
pared to the real diameter of the phantom used (e.g. 240 mm for the body phantom). The
diameter depends on the gradient sensitivity (see (Gradient Sensitivity / p. 34)).
Intensity profile
The intensity profile (uniformity) refers to the ability of the MR-imaging system to produce
a constant signal response throughout the scanned volume in case a homogeneous
phantom is used.
To minimize the effect of noise in the image, it is convolved with a nine point low-pase-fil-
ter function. The uniformity “U“ is calculated over 80% of the Diameter in the horizontal
and vertical direction:
U = (Smax - Smin) / (Smax + Smin) × 100%
S = signal level

MAGNETOM C! M4-030.850.20.01.02 Page 40 of 50 Siemens AG

09.06 SMMR Medical Solutions
Tune-up and QA 41

Fig. 36: DIP Results

General QA 0

Image Orientation
The image orientation test ensures that the MR-images are oriented properly and that the
axes are labeled correctly.

NOTE This test is the most important measurement. Make sure, image
orientation is always correct!

Measurement
The image orientation measures two images:
• Transverse orientation TR and TE are very short, so that the small bottle can be seen to
the right of the large spherical phantom.

Fig. 37: 1st Measurement

Siemens AG M4-030.850.20.01.02 Page 41 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
42 Tune-up and QA

• Oblique slice (TRA tilted 45 into SAG)

The evaluation tries to find the filling connector to the right on top of the large spherical
phantom.

Fig. 38: 2nd Measurement

RF-Verify
This measurement is the same as the last part of the RF-characteristic measurement from
the tune-up. (see (RF-Characteristics / p. 13))

Linearity Check
This measurement is the same as the Gradient Linearity from the gradient test tools.

NOTE This test involves the whole system!

MAGNETOM C! M4-030.850.20.01.02 Page 42 of 50 Siemens AG

09.06 SMMR Medical Solutions
Tune-up and QA 43

Fig. 39: Example of the Y-Gradient

Shim Check
This measurement uses the Phantom Shim measurement from the tune-up without cor-
rection of the gradient offset values. (see (Phantom Shim / p. 21))

ECC Check
This measurement uses the ECC measurement from the tune-up without correction of
the ECC-values. (see (Eddy Current Compensation / p. 25))

Grad Delay Check

This measurement uses the Grad Delay measurement from the tune-up without correc-
tion of the delay times. (see (Gradient Delay / p. 32))

Calc Artifacts
Phase-encoding direction
With “Calcar“ (Calculate artifact) the amount of ghosts/artifacts outside the phantom in
phase-encoding direction is analyzed. We are mainly looking for multiple blurred ghosts
caused by instable system components.
Parameters
The CALCAR measurement uses a double-echo sequence with the following fixed
parameters:

TR=600ms TE=35/105 ms Thickness=10 mm

FOV=500 mm MA=256x256 AC=1

Evaluation

Siemens AG M4-030.850.20.01.02 Page 43 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
44 Tune-up and QA

The evaluation checks if the signal-to-noise level is sufficiently high and a circular phan-
tom is used.
The mean value is calculated in square ROI's placed in the corners of the image (= Sig-
nalNoise) and in a square ROI in the center of the object (= SignalObject).
Along the phase-encoding direction in the area corresponding to the width of the object,
the average is taken from every 5 × 5 pixel-block of the background. The maximum value
(= Signalartifact) determines the amount of artifact as a relative percentage with respect to
phantom intensity.
The results are compared with the specified values and a message is printed out telling
you if the system meets specifications.
Artifact[%] = (Signalartifact - SignalNoise) / SignalObject × 100

MAGNETOM C! M4-030.850.20.01.02 Page 44 of 50 Siemens AG

09.06 SMMR Medical Solutions
Tune-up and QA 45

Fig. 40: Calcar 1st Echo

Spike Check
General
Spikes are signals detected by the receiving coil which result in a wave-like pattern (moire
structure). The influence to the image depends on the amount, position, and intensity of
the received spikes.

Siemens AG M4-030.850.20.01.02 Page 45 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
46 Tune-up and QA

Stability Check
The MAGNETOM C! is quite sensitive to external time-varying magnetic fields caused by
moving magnetic objects such as trains, trams, cars, elevators, wheelchairs, etc. or fields
of time-varying currents in the vicinity of the magnet.
Two sequences
The effect of these different external field sources on the magnetic field of the MR-system
are checked by means of two sequences:
• Spin echo sequence checks for periodic line voltage interferences (162/3Hz, 50 or 60
Hz)
• Gradient echo sequence (Flash 2D) checks for low frequency or statistical magnetic
field (Bo) instabilities caused by e.g. moving objects.
No phase-encoding gradient
The measurements are done without phase encoding gradient producing equal raw data
lines if there are no disturbances. The differences that occur are evaluated and compared
to specifications.
EFI unit
If the stability check is not in specification due to external field interferences, the so-called
EFI unit has to be installed.
Additional, the “Easy Site“ flag in the configuration of the MR-system can be set to “No“.
This limits the maximum possible echo time TE for gradient echo sequences to <25 ms
and all protocols using longer TEs will be dimmed.

Synthesizer
WIP

Field Stability
WIP

MAGNETOM C! M4-030.850.20.01.02 Page 46 of 50 Siemens AG

09.06 SMMR Medical Solutions
Changes 47
3-
2Changes

In order to get a quick overview what has changed, here you find the changes to the last
version.

Siemens AG M4-030.850.20.01.02 Page 47 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
48 Changes

Changes from V1.0 to V2.0 0

Part ‘Prefix’ 0

Where What is changed?

(M4-030.850.10 / Change the descriptions of “What are the Disadvantages?“
What are the Disad-
vantages?)

Part ‘System’ 0

Where What is changed?

(M4-030.850.10 / Update the caution.
Overview)
(M4-030.850.10 / Update the figure “Electronic Cabinet“
Electronic Cabinet)

Part ‘Software’ 0

Where What is changed?

(M4-030.850.11 / Delete the description of HKEY_LOCAL_MACHINE.
syngo MR)
(M4-030.850.11 / Delete the following section: IP Addressing.
DICOM)
(M4-030.850.11 / Net- Delete the graphic of Service Access Manager II.
work)
(M4-030.850.11 / Ser- Delete the following section: Remote diagnostic (RDIAG)
vice Software)
(M4-030.850.11 / Sie- Siemens Remote Services (SRS) added
mens Remote Ser-
vices (SRS))
(M4-030.850.11 / Update the figure “Software installation 1“
Software Installation)
(M4-030.850.11 / Update the figure “Software installation 2“
Installation Proce-
dure)
(M4-030.850.11 / Update the figure “Software configuration“
Configuration)

MAGNETOM C! M4-030.850.20.01.02 Page 48 of 50 Siemens AG

09.06 SMMR Medical Solutions
Changes 49

Part ‘Host/Imager(MRIR)’ 0

Where What is changed?

(M4-030.850.12 / Figure “Host“ added
Configuration)

Part ‘Control’ 0

Where What is changed?

(M4-030.850.13 / Update the figure “MC4C40/GCTX Block Diagram“
GCTX)

Part ‘RF System’ 0

Where What is changed?

(M4-030.850.14 / Update the figure “RF-Components in ECA“
Overview)

Part ‘Power Distribution’ 0

Where What is changed?

(M4-030.850.19 / Update the figure “LPD components“
Overview)
(M4-030.850.19 / Update the figure “Power-up circuit“
Power-up Circuit)
(M4-030.850.19 / Switch on Timer added
Power-up Circuit)

Part ‘Tune-up and QA’ 0

Where What is changed?

(Tune-up Procedures Update the figure “Tuneup“
only with
Body-Spine-Array
large (since MR
2002B) / p. 5)

Siemens AG M4-030.850.20.01.02 Page 49 of 50 MAGNETOM C!

Medical Solutions 09.06 SMMR
50 Changes

MAGNETOM C! M4-030.850.20.01.02 Page 50 of 50 Siemens AG

09.06 SMMR Medical Solutions