Acoustics 08 Paris

Don’t forget the quench pipe when installing an MRI

C. Ostendorf
Cauberg-Huygen R.I. BV, Postbus 480, 6200 AL Maastricht, Netherlands
[email protected]

5357
Acoustics 08 Paris

An MRI is a useful medical device but it makes a lot of noise. A sound level of 90 dB(A) or more in the MRI
room is not unusual. Placing an MRI into an existing hospital means that extra care has to be taken to prevent the
noise from the MRI causing nuisance in adjacent rooms. In this paper the situation is discussed in which
complaints appeared after the installation of the MRI. First sound measurements were done to see if the sound
level in the office above the MRI fulfils the noise ratings. This way it would be clearer if the acoustic measures
did their jobs. Whatever the outcome, the hospital wanted to put an end to the complaints. So, more sound
measurements were done to establish the cause of the nuisance and to point out what extra measures had to be
taken to solve the problem. It turned out that the quench pipe played an important role in this situation.

2.2 Result overall level

1 Introduction
Figure 1 shows the result for the A-weighted overall level
In 2004, a 3T (tesla) MRI was placed on the 4th floor in the during a CASL sequence. The top graph shows the sound
Leiden University Medical Centre. During the construction level in the machine room, the bottom graph the sound level
of the MRI room, extra acoustic measures were taken to in the office above the MRI room. Every slice in the graph
prevent nuisance from the noise of the MRI. The sound gives the LAeq over 1 second.
level in the rooms above the MRI should not exceed 40
dB(A). Nevertheless, complaints appeared from the rooms
on the floor above the MRI (5th floor) when the MRI was
performing a CASL sequence. The CASL sequence
sounded like a phone ringing but no one could find the
phone.
First sound measurements were done to see if the sound
level in the room above the MRI meets the noise ratings.
This way it would be clearer if the acoustic measures were
doing their job. Whatever the outcome, the hospital wanted
to put an end to the complaints. So, more sound
measurements were done to establish the cause of the
nuisance and to point out what extra measures had to be
taken to solve the problem.

2 Sound measurements
Fig. 1 Profile of A-weighted sound level during CASL
sequence. Top graph shows sound level close to MRI,
2.1 General bottom graph the sound level in the office.

In order to make sure the right sound level was reviewed, The effect of the CASL sequence is clearly visible in the
two microphones were used. One microphone, type B&K top graph. Every 2 seconds a 2 second noise pulse is given.
4165 was placed in a small machine room next to the MRI The difference between the situation with and without noise
room. In the machine room there was a small gap in the is about 30 dB(A). In the office, the background sound
wall to the MRI room for cables. The microphone was level is 34.5 dB(A). The CASL sequence produces about 40
placed in front of the gap. With this microphone changes in dB(A).
sound level due to the sequence of the MRI, could be The averaged sound level of a complete CASL sequence is
measured clearly. 36.2 dB(A) in the office. This is only 2 dB(A) above the
The second microphone, also type B&K 4165, was placed background level. When corrected for the background level,
in the room directly above the MRI. This room was used as the sound level of the MRI is 31 dB(A) which fulfils the
a small office and for study. The microphone was placed at required 40 dB(A).
a height of 1.6 m above the ground. Nevertheless, the noise of the MRI is very audible. This is
Both microphones were connected to a two channel real caused by the specific frequencies of the CASL sequence.
time frequency analyzer type 2144 from B&K. So, the Figure 2 shows the frequency analysis in 1/3rd octave bands
sound measurements were done simultaneously for both of the period the CASL sequence is producing sound.
locations.
Every second an averaged sound level Leq was logged so it
was possible to view the profile of the noise in time. At the
same time a CPB frequency analysis was made in 1/3rd
octave bands.

5358
 Acoustics 08 Paris

Sound spectrum of CASL sequence superconducting capabilities of the electromagnetic coils

90.0 within the MRI. The most commonly used coolant is
80.0
CASL
helium. In case of emergency, the magnet of the MRI will
70.0
be shut down. This causes the helium to boil and change
60.0
from liquid into gas very rapidly. The expanding gas should
50.0
leave the machine and building as soon as possible and for
Leq [dB]

40.0
this the quench pipe is used. The quench pipe is a safety
30.0
device. Figure 4 shows a picture of the situation.
20.0

10.0

0.0

00

50

00

00

00

00

0
0

50

00

00
25

.5

40

50

63

80

00

00
80
10

12

16

20

25

31

40

50

63
31

50

63

80
10

12

16

20

25

31

40

10
Frequency [Hz 1/3 octave band]

Fig. 2 sound spectrum in 1/3 octave band of CASL office

sequence

Figure 2 shows that the 1250 Hz is very prominent in the

sound spectrum. Therefore in figure 3 the profile of the
1250 Hz band is given.
quench pipe

Fig. 4 Situation of quench pipe and office

From this situation it was clear that airborne noise was a

factor in the noise problem. Construction noise could be
involved as well because the quench pipe inside the
building was attached to the building construction which
was in this case the floor of the office. Figure 5 shows this
in detail.

Fig. 3 Profile of 1250 Hz sound level near MRI (top graph)

and in office above MRI (bottom graph).

Now it is clear why the noise of the CASL sequence can be

heard so clearly. The difference between background level
and CASL sound level is about 12 to 18 dB in the office.

3 Cause of the sound level

3.1 General
In general three causes are possible for the sound level in Fig. 5 Construction detail of quench pipe connected to the
the office: building.
1. Airborne noise from the MRI room to the office;
2. Construction noise from equipment attached to the
MRI and attached to the building construction; 3.2 Airborne or construction
3. Combination of both.
By listening to the sound of the CASL sequence in the In order to determine whether the noise in the office was
office, it appeared that the sound was everywhere although caused by airborne noise, construction noise or a
close to the window it could be heard better. After opening combination of both, the noise of the quench pipe was
the window, it became quite clear that the main sound reduced. If airborne noise was the only cause, the reduction
source was outside the room. A small pipe on the roof in sound level for 1250 Hz close to the quench pipe would
terrace was the outlet of the quench pipe. In order to be the same as the reduction of sound level in the room. If
operate, the MRI uses a coolant to enable the construction noise is the only cause, reducing the airborne

5359
Acoustics 08 Paris

noise of the quench pipe would not have any effect on the Sound level office 5.15

sound level in the office at all. 55

The sound reduction of the quench pipe was achieved by 50 office 5.15 background noise
office 5.15 noise level incl CASL

placing an enclosure around the quench pipe. The enclosure 45 office 5.15 noise CASL reduced

was made of sound absorbing panels which happened to be 40

stored in the hospital because they were not used at that 35

Leq [dB]
time. Since the quench pipe is a safety device, it was not 30

allowed to build a rigid construction. If an explosion would 25

occur during the sound test, the sound enclosure had to be 20

easily blown away. So the panels were just placed against 15

each other and gaps were closed with extra sound absorbing 10

material. Figure 6 shows the result.

00

00

00

50

00

00

00

00

0
0

00

50
0

0
25

.5

40

50

63

80

00
63

80
10

12

16

20

25

31

40

50
31

80
31

40

50

63
10

12

16

20

25

10
Frequency [Hz 1/3 octave]

Fig. 8 Sound level in the office with and without enclosure.

Without the enclosure, the sound level of 1250 Hz is an

averaged 10 dB over the background level in the office.
With the enclosure the sound level of the CASL sequence is
reduced by 5 dB. This is not as much as the reduction close
to the quench pipe. This means that construction noise has
some influence but that airborne noise is the most important
factor. So, the first measure is to reduce the noise of the
quench pipe. If after those measures the noise of the CASL
sequence is still a nuisance, then measures to reduce the
influence of constructing noise have to be taken.

Fig. 6 Sound enclosure quench pipe.

4 Sound reducing measures

To reduce the noise of the quench pipe, four simple

Before and after the enclosure was placed, a sound
measures can be taken:
measurement was done at approximately 1 meter distance
of the quench pipe. The result of the measurement with and 1. put a silencer in the outlet of the quench pipe;
without enclosure is shown in figure 7. 2. put the quench pipe in a enclosure which isolates
Sound level close to quench pipe
sound well;
70 3. replace the glass of the windows of the offices
65 Q-pipe without reduction
Q-pipe reduced
with better isolating glass;
60

55 4. change the setup of the CASL sequence.

50

45 Due to safety regulations it was not allowed to build a rigid

Leq [dB]

40 enclosure with sound dampened outlets. Putting a silencer

35

30
in the quench pipe wasn’t allowed either because of these
25
regulations. The width of the quench pipe is carefully
20 chosen and related to the length of the pipe, the amount of
15
coolant gas and the time needed to let the coolant gas
10
escape. Changing the length or diameter of the pipe would
00

00

00

0
0

00

50

00

00

00

50

00
25

.5

40

50

63

80

00
80
10

12

16

20

25

31

40

50

63
31

40

50

63

80
10

12

16

20

25

31

10

Frequency [Hz 1/3 octave] change the reaction of the quench pipe in case of an
Fig. 7 Sound spectrum of quench pipe (distance 1 m) with emergency. Furthermore, welding things to the pipe would
and without reducing enclosure. cause weak spots and potential danger.
Replacing the glass in the windows is a simple but effective
Figure 7 shows the relevance of the 1250 Hz band. The solution. To be able to place the right acoustic glass, it
temporary enclosure has achieved a reduction of 7 dB in turned out that most of the façade would have to be
this frequency band. If only airborne noise is the cause of replaced. Because of the number of windows that had to be
the problem, the sound level in the office will be reduced treated this way, this solution was too expensive. Besides,
with 7 dB as well for the 1250 Hz band. If the reduction is users of the offices wanted to be able to open their windows
less, construction noise has an influence. or doors without hearing too much noise of the pipe. So
reducing the noise of the quench pipe was strongly
In figure 8 the result is given for the influence of the preferred.
enclosure on the sound level in the office. Three lines are
shown in the graph: the background level if the CASL The MRI technicians of the hospital were able to change
sequence is not active, the sound level with active CASL the setup of the CASL sequence thus producing fewer
and the sound level with enclosure around the quench pipe. specific tones and less noise.

5360
 Acoustics 08 Paris

5 Conclusion

Only by changing the setup of the CASL sequence, it

turned out to be possible to solve the noise problem of the
MRI. This solution means that the use of the MRI is limited
instead of using technical measures to prevent the noise
from causing nuisance. For a research hospital, this is not a
desired situation. The MRI department wants to do
whatever is necessary without having to be concerned about
the neighbours.
In this case, technical measures to reduce the noise of the
MRI after installation of the MRI could only be taken
against high costs. During the engineering process of the
MRI installation, the outlet of the quench pipe has to be
taken into account. This increases the chances of unlimited
use of the MRI and decreases the chances of nuisance.
Everybody happy!

Acknowledgments
Thanks to the Leiden University Medical Centre for their
support.

5361