Kidney motion during free breathing and breath hold for MR-guided radiotherapy

This article has been downloaded from IOPscience. Please scroll down to see the full text article.

2013 Phys. Med. Biol. 58 2235

(http://iopscience.iop.org/0031-9155/58/7/2235)

Download details:

IP Address: 144.173.6.37

The article was downloaded on 16/04/2013 at 09:06

Please note that terms and conditions apply.

View the table of contents for this issue, or go to the journal homepage for more

Home Search Collections Journals About Contact us My IOPscience

IOP PUBLISHING PHYSICS IN MEDICINE AND BIOLOGY

Phys. Med. Biol. 58 (2013) 2235–2245 doi:10.1088/0031-9155/58/7/2235

Kidney motion during free breathing and breath holdfor MR-guided radiotherapy

Mette K Stam1, Marco van Vulpen1, Maurits M Barendrecht2,Bernard A Zonnenberg3, Martijn Intven1, Sjoerd P M Crijns1,Jan J W Lagendijk1 and Bas W Raaymakers1

1 Department of Radiotherapy, University Medical Center Utrecht, Heidelberglaan 100, 3584 CX,Utrecht, The Netherlands2 Department of Urology, University Medical Center Utrecht, Heidelberglaan 100, 3584 CX,Utrecht, The Netherlands3 Department of Internal Medicine, University Medical Center Utrecht, Heidelberglaan 100,3584 CX, Utrecht, The Netherlands

E-mail: M.K.Stam-3@umcutrecht.nl

Received 8 January 2013, in final form 12 February 2013Published 11 March 2013Online at stacks.iop.org/PMB/58/2235

AbstractCurrent treatments for renal cell carcinoma have a high complication rate due tothe invasiveness of the treatment. With the MRI-linac it may be possible to treatrenal tumours non-invasively with high-precision radiotherapy. This is expectedto reduce complications. To deliver a static dose distribution, radiation gatingwill be used. In this study the reproducibility and efficiency of free breathinggating and a breath hold treatment of the kidney was investigated. For 15patients with a renal lesion the kidney motion during 2 min of free breathingand 10 consecutive expiration breath holds was studied with 2D cine MRI.The variability in kidney expiration position and treatment efficiency for gatingwindows of 1 to 20 mm was measured for both breathing patterns. Additionallythe time trend in free breathing and the variation in expiration breath holdkidney position with baseline shift correction was determined. In 80% of thepatients the variation in expiration position during free breathing is smallerthan 2 mm. No clinically relevant time trends were detected. The variation inexpiration breath hold is for all patients larger than the free breathing expirationvariation. Gating on free breathing is, for gating windows of 1 to 5 mm moreefficient than breath hold without baseline correction. When applying a baselinecorrection to the breath hold it increases the treatment efficiency. The kidneyposition is more reproducible in expiration free breathing than non-guidedexpiration breath hold. For small gating windows it is also more time efficient.

0031-9155/13/072235+11$33.00 © 2013 Institute of Physics and Engineering in Medicine Printed in the UK & the USA 2235

2236 M K Stam et al

Since free breathing also seems more comfortable for the patients it is thepreferred breathing pattern for MRI-Linac treatments of the kidney.

(Some figures may appear in colour only in the online journal)

1. Introduction

The current standard of care for T1 staged renal cell carcinoma (RCC) is a partial nephrectomy.This intervention has a high complication rate and associated morbidity mainly caused bybleeding and urine leakage (Gill et al 2003). Alternative treatments, such as cryo ablation(Wyler et al 2007) or radio-frequency (RF) ablation (Park et al 2006), also have complicationsdue to the invasiveness of the treatment. The most common complications are bleedings andhaematuria for RF-ablation and nerve and endothelial injury for cryo ablation (Atwell et al2012). A non-invasive treatment of RCC, such as external beam radiotherapy, is expected toimprove the complication rate.

Currently, radiotherapy for kidneys is limited due to the uncertainties in tumour positioncaused by respiratory motion (Moerland et al 1994). With recent advances in image-guidedradiotherapy, such as the Cyberknife (Adler et al 1997), cone beam computed tomography(Jaffray et al 2002) and MRI-Linac (MRL) (Lagendijk et al 2008), it becomes possible todetermine the tumour and organs at risk positions non-invasively and compensate for motion.Possible methods to compensate respiratory motion are tumour tracking and gating (Kubo andHill 1996, Keall et al 2001). Tracking is the most time efficient solution with a continuousradiation delivery, but it is technically and dosimetrically the most challenging option sincethe radiation beam has to follow the tumour. Therefore, for the first kidney treatments on theMRL we consider a gated radiation delivery.

At our institution an MRL is under development in cooperation with Elekta, Sweden andPhilips, The Netherlands (Raaymakers et al 2009). This is a combination of a 6 MV linac and1.5 T MR-scanner. Also other institutions are developing an integrated MRI and teletherapyunit (Fallone et al 2009, Dempsey et al 2005). The technical feasibility of both gating andtracking on our MRL prototype was shown in phantom studies (Crijns et al 2011, 2012). TheMRL will provide MR-imaging during irradiation. For the kidney fast motion detection usingMR-navigators has been shown previously (Stam et al 2012). Furthermore, the feasibility ofdelivering 25 Gy to the renal tumour in an MRL was previously studied under the assumptionof a static kidney position (Kerkhof et al 2011), i.e. when a motion compensation strategy,such as gating, is applied.

In order to bring an MRL treatment of the kidney into the clinic an optimal breathingstrategy for the treatment should be found for a gated radiation delivery. There are severaloptions for free breathing and breath hold. Especially for the latter guidance methods such asactive breathing control (ABC) or audiovisual feedback are used (Wong et al 1999, Vedamet al 2001). The results of Korreman et al (2005) for non-guided breath hold in breast cancerpatients are, as they point out, comparable to the result of Remouchamps et al (2003) whoused ABC. This suggests that the reproducibility of breath hold organ positions is comparablefor non-guided and guided or coached breath hold. Also, in our clinic the majority of thepatients with left sided breast tumours are treated under non-guided voluntary breath hold.The idea behind the MRL is to accurately monitor the tumour position with real-time imagingand adapt the treatment to the patient, making it a very precise treatment. Therefore, we wouldlike to quantify kidney movements and investigate the use of voluntary breath hold versus

Kidney motion during free breathing and breath hold for MR-guided radiotherapy 2237

free breathing gating for an MRL treatment. We studied the stability and reproducibility ofthe kidney position during uncoached free breathing and non-guided breath hold in patientswith a renal lesion using cine MRI. Furthermore, the treatment efficiency for a gated radiationdelivery during both breathing patterns was determined.

2. Materials and methods

Fifteen patients were included in this study, which is approved by our medical-ethical hospitalreview board. From all patients we obtained written informed consent. All patients were eitherunder active surveillance or awaiting surgical treatment for their renal lesions. The inclusioncriteria for the study were a renal lesion and no contra-indications for MRI. Therefore, thepatient demographics in this study are representative of the patient group treated in ourhospital. However, since our institution is specialized in treating patients with Von HippelLindau syndrome a relatively high percentage of the patients (33%) developed the renal lesionwithin this syndrome.

All patients were asked to perform free breathing as well as breath hold in the MR-scannerwhile we imaged the kidney dynamically, resulting in 2D movies of the kidney motion. Forfree breathing the patients were instructed to breath normally and for the breath hold to holdtheir breath during expiration as long as it was comfortable and to resume breathing afterwards.The breath hold scans were stopped after capturing one complete breathing cycle followingthe breath hold period. Before breath hold the patients performed two coached deep breaths.Before the start of the scan the patients performed one breath hold as training.

2.1. Image acquisition and registration

All MR-imaging was done on an 1.5 T Philips Achieva system using an eight-channel bodyreceive array. Both breathing scenarios were examined with coronal as well as sagittal 2Ddynamic sequences. The 2D coronal balanced steady state free precession (bSSFP) scans hadTE = 1.4 ms, TR = 2.8 ms, NSA = 1, voxel size = 2.0 × 2.0 mm, slice thickness = 7 mm,FOV = 320 × 250 mm SENSE = 2 and dynamic scan time = 0.44 s. The sagittal imageshad the same parameters, except for TE = 1.2 ms, TR = 2.4 ms, FOV = 450 × 363 mm anddynamic scan time = 0.40 s. The coronal scans were angulated in the direction of motion ofthe kidney as estimated on a dynamic 2D sagittal scan of the kidney during free breathing.The duration of the free breathing scans was 121 s for the coronal acquisition and 124 s forthe sagittal acquisition.

All analyses were performed offline. The time-frames of the 2D movies were registered tothe sixth frame of that dynamic scan using the Elastix toolbox (Klein et al 2010). In the sixthframe the contrast of the steady state of the bSSFP sequence was established. A delineationof the kidney was made on this reference frame only and, with a 15 mm margin, used as theregion of interest for the rigid registration. The other time frames were not delineated. Unlessstated otherwise, the Euclidean distance, calculated form the registration results, was used forthe analyses. The distances were taken positive if the kidney was superior to the referenceposition and negative if inferior.

2.2. Free breathing

A sagittal dynamic scan of the kidney was acquired during free breathing. After that a coronaldynamic scan was acquired, also during free breathing. In both scans the expiration peaks weredetermined and the mean and the standard deviation of those peak positions were calculated

2238 M K Stam et al

as these indicate the reproducibility of the kidney position during expiration. A linear fit ofthe expiration kidney positions was made as well. The slope of the fit indicates whether thereis a time trend.

To determine the efficiency of a free breathing gated treatment the percentage of timein which the kidney was inside a gating window was calculated. This gating window waspositioned symmetrically around the average expiration position of the kidney in the firstfive breathing cycles. The calculation was done for multiple gating windows varying in sizebetween 1 and 20 mm.

2.3. Breath hold

All patients performed ten consecutive expiration breath holds in total, which were toleratedwell. Coronal scans were made of the first five breath holds and sagittal images of the lastfive breath holds. The position of both imaging planes was exactly the same as during thefree breathing scans. This included the angulation of the coronal plane based on the observedkidney motion in the free breathing sagittal scans. The stability and the reproducibility ofthe breath-hold were measured. The drift on the kidney position as a function of breath holdduration indicates the stability. The reproducibility of the breath hold kidney position is givenby the change in centre of gravity of the delineations of the kidney, which were made togenerate the ROI that aided the registration (see section 2.1).

To calculate the treatment efficiency for breath hold, an acceptance window (the sameas a gating window for free breathing) was positioned on the reference kidney position, i.e.the sixth frame, in the first coronal breath hold. It was positioned symmetrically around thereference position and the size was varied between 1 and 20 mm. The percentage of time inwhich the kidney was inside the acceptance window was determined.

We also investigated whether correcting the position of the acceptance window forbaseline shifts between breath holds would improve the treatment efficiency. This was doneby repositioning the acceptance window on the kidney position in the sixth frame of everybreath hold scan and calculating the percentage of time in which the kidney was inside theacceptance window. Also for this calculation the size of the window varied between 1 and20 mm.

3. Results

The 2D cine MR images are acquired with a good contrast of the kidney. They are shown infigures 1(a) and (b). The acquisition time and resolution are as given in section 2.1

3.1. Free breathing motion of the kidney

The variation in kidney position, during free breathing, along the cranial-caudal, anterior–posterior and right–left axes are given in figure 2. It is clear that the largest component of thekidney motion is in the cranial-caudal direction (figure 2(a)). The motion also has a smallercomponent in the anterior–posterior direction. The smallest component is in the right–leftdirection (figure 2(c)), where five patients (33%) had a variation smaller than 1 mm over allbreathing cycles in the dynamic scan. In only four patients (27%) a displacement lager then2 mm was measured during the 124 s of scanning.

The variation in expiration positions during free breathing for all patients is shown infigure 3(a). These results show that this position during free breathing is in 80% of the patientsreproducible within 2 mm. Only 20% has a variation in expiration kidney position of more

Kidney motion during free breathing and breath hold for MR-guided radiotherapy 2239

(a) (b)

Figure 1. (a) Coronal and (b) sagittal time frame in 2D cine MRI.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Patient #

− 30

− 20

− 10

0

10

Dip

lace

men

t fro

m m

ean

[mm

]

Variation in CC position free breathing

(a)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Patient #

− 10

− 5

0

5

Dip

lace

men

t fro

m m

ean

[mm

]

Variation in AP position free breathing

(b)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Patient #

− 6

− 4

− 2

0

2

4

Dip

lace

men

t fro

m m

ean

[mm

]

Variation in RL position free breathing

(c)

Figure 2. The variation in kidney position during free breathing with respect to the (a) cranial-caudal(CC), (B) anterior–posterior (AP), and (c) right–left (RL) axis. The CC and AP displacements weremeasured in the sagittal dynamic scan and the RL displacements in the (angulated) coronal scan.NB: all plots have a different y-scale.

2240 M K Stam et al

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Patient #

− 10

− 8

− 6

− 4

− 2

0

2

4

6

8

Var

iatio

n in

Exp

iratio

n po

sitio

ns [m

m]

The stability of the expiration positions during free breathing

(a)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Patient #

− 10

− 8

− 6

− 4

− 2

0

2

4

6

8

Dis

plac

emen

t [m

m]

Variation in kidney position between breath holds

(b)

Figure 3. (a) The differences in expiration position during free breathing with respect to the averageexpiration position during the first five expirations. (b) Boxplot of the breath-hold kidney positionat the breath hold start.

than 2 mm while for 53% of the patients it is even smaller than 1 mm. The fits through theexpirations show an average time trend for the entire patient group of 0.2 ± 0.4 mm min−1,which means that on group-level no significant time trend could be detected. We consider thetime trend significant if the value 0 mm min−1, which represents a straight line, falls outsidethe 95% confidence interval of the fit. For some individual patients there was a statisticallysignificant time trend. All these time trends per minute are smaller than the variation on theexpiration positions. Therefore we do not consider these time trends clinically relevant.

Figure 4(a) shows the gating efficiency for free breathing. As expected, the efficiency isincreased with an increased gating window. There is a lot of variation in efficiency betweenpatients.

3.2. Breath hold position variations of the kidney

As an example, the ten breath holds of patient 4 are plotted consecutively in figure 5. Thisimage shows that there are two effects that influence breath hold stability: variation in startingposition of the breath holds and drift during the breath hold. The variation in starting position ofthe breath holds is shown in figure 3(b). Furthermore, the percentages of the breath holds thatfall within the acceptance window are shown in figure 4. Figure 4(b) shows the percentagesif the acceptance window is positioned on the reference frame of the first breath hold andused for all following breath holds. We also calculated the efficiency when the acceptancewindow is repositioned at the start of each breath hold, see figure 4(c). The data show thatthis improves the treatment efficiency especially for the smaller gating/acceptance windows.In comparison with the free breathing efficiency (figure 4(d)) for the windows smaller than5 mm the uncorrected breath hold is less efficient and the corrected breath hold more efficient.

The given percentages are merely the percentage of the breath holds in which the kidneywas within the acceptance window. The time needed to prepare the patient for the breath holdis not considered. In our study the time between breath holds (the breath hold preparationtime) was on average 43 s. It is hard to predict how the treatment efficiency will be influencedby the breath hold preparation since this might coincide with gantry movements. Peng et al(2011) found that with breaks of 15 s the duty cycle due to breath hold duration was in practicebetween 0.6 and 1, depending on the number of breath holds needed to deliver a field. If wewould have the same results, the uncorrected breath hold treatment efficiency will likely beless than the free breathing gating efficiency for all acceptance/gating window sizes.

Kidney motion during free breathing and breath hold for MR-guided radiotherapy 2241

0 5 10 15 200

10

20

30

40

50

60

70

80

90

100

Gating window [mm]

Per

cent

age

of ti

me

kidn

ey is

with

in g

atin

g w

indo

w [%

]Free breathing

(a)

0 5 10 15 200

20

40

60

80

100

Acceptance window [mm]Per

cent

age

of ti

me

kidn

ey is

with

in g

atin

g w

indo

w [ % Breath hold without baseline correction

(b)

0 5 10 15 200

20

40

60

80

100

Acceptance window [mm]Per

cent

age

of ti

me

kidn

ey is

with

in g

atin

g w

indo

w [% Breath hold with baseline correction

(c)

0 5 10 15 20 250

20

40

60

80

100

120

Gating / acceptance window [mm]

Gat

ing

effic

ienc

y [%

]

Average gating efficiency during imaged breathing pattern

breath hold

baseline corrected breath hold

free breathing

(d)

[

[

Figure 4. The percentage of time the kidney is within the gating or acceptance window as a functionof the size of the window for (a) free breathing, (b) uncorrected breath hold and (c) breath hold withbaseline shift correction. (d) is the population average for all three breathing strategies. It is clearthat for the smaller gating or acceptance windows free breathing is more efficient than uncorrectedbreath hold, but baseline corrected breath hold is more efficient than free breathing. However, thebreath hold preparation time is not considered in any of these analyses.

4. Discussion

4.1. Free breathing variations

For free breathing the kidney moves predominantly in the cranial-caudal direction, with asmaller component in the anterior–posterior direction. The kidney excursions in the right-leftdirection are smaller than 2 mm for 11 patients (73%). This is in line with the observationof Moerland et al (1994) that the movement of the kidney predominantly occurs in a tiltedcoronal plane through the longitudinal axis of the kidney. The expiration positions are for mostpatients reproducible within 2 mm. The intrafraction variation in exhale position of the liverfound by Case et al (2009) are larger than we found for the kidney. However, when consideringthat respiration induced liver motion is about 1.5 times larger than respiration induced kidneymotion (Bussels et al 2003) the average variation of 3.0 mm found by Case et al (2009) seemscomparable to our results.

2242 M K Stam et al

−25

−20

−15

−10

−5

0

5

10

Time −>

Dis

plac

emen

t [m

m]

Breath holds patient 4

Figure 5. The ten consecutive breath holds of patient 4.

For a single fraction treatment 10 min beam-on time on a conventional accelerator isneeded (Kerkhof et al 2011). If a time trend would be present the expirations might falloutside the gating window which will reduce the gating efficiency during the treatment. Inour study we only measured over a period of 2 min in which no relevant time trend wasdetected. These results are in line with Nøttrup et al (2007), who found only a significanttime trend in one of the 11 lung cancer patients they analysed over 100 s. However, to drawdefinite conclusions a longer measurement period is necessary. Still, the results of Case et al(2009), for the liver, were obtained by measuring the expiration position only before and aftertreatment delivery leading to a median time between the measurements of 12 min. In thistime they found no significant correlation between treatment time and intrafraction change intreatment position. Therefore it seems likely that also for the kidney this correlation is notpresent.

4.2. Breath hold variations

The results of the breath hold reproducibility are comparable to the results of Dawson et al(2001) for liver with ABC and Peng et al (2011) for lung tumours with visual feedback.Schwartz et al (1994) found a mean deviation in expiration breath hold position of the kidneyof 2.8–3.6 mm and a maximum range of 0–6.8 mm. Although we have a few larger outliers, theresults are in the same range. The variability of expiration breath hold position is bigger than thevariability of expiration position during free breathing for all individual patients. Korremanet al (2005) found the same for breast cancer patients. The difference is also clear in thepercentages of time the kidney is within the acceptance window (see figures 4(b) and (a)). Forthe small windows, i.e. 1–5 mm, the treatment efficiency is higher during free breathing thanduring breath hold. For a precise treatment it is important to use a small gating/acceptancewindow to prevent blurring of the steep dose gradients at the edge of the planning targetvolume. For a 3 mm gating window the fraction of under dosage is clearly smaller than for agating window of 12.5 mm according to Dietrich et al (2005). The dose blurring as a functionof gating window for the kidney will be future work. Furthermore, a baseline correction atthe start of each breath hold does improve the efficiency, making it more efficient than freebreathing. However, this will need multiple on-line plan adaptations making the treatmentmore complex and is therefore not preferred.

Kidney motion during free breathing and breath hold for MR-guided radiotherapy 2243

In this comparison the breath hold preparation time was not considered. As argued insection 3.2 it is hard to predict the required additional time for breath hold preparationduring treatment. Nonetheless, it seems reasonable to expect a significant prolongation ofthe treatment, which will make the beam on time percentage worse for breath hold than forfree breathing. There are different methods, such as ABC (Dawson et al 2001, Eccles et al2006) and audiovisual feedback (Berson et al 2004, Linthout et al 2009), available that mightimprove the breath hold stability and reproducibility. However, for patient comfort, breathhold is not desired. Although the patients in this study tolerated the breath holds well, it hasno doubt that free breathing is more comfortable for the patients, especially for patients withimpaired lung function, obesity and generally elderly patients. Since the treatment efficiencyof the uncoached free breathing is also better in this study than the uncoached breath hold,free breathing gating is preferred for the MRL treatments.

4.3. Improving free breathing stability

An option to improve the gating efficiency during free breathing is audiovisual feedback(Vedam et al 2001). This improves the reproducibility of the breathing pattern (Kini et al2003). It is also used in diagnostic MRI to improve respiratory gated acquisitions. A navigatorsignal can be used as input for visual guidance (Jhooti et al 2011). Since navigators will beused for fast kidney position monitoring in the MRL (Stam et al 2012), audiovisual feedbackcould be used to improve free breathing gating efficiency, making free breathing gating evenmore suitable for kidney radiotherapy treatments. This will be investigated for the clinicalintroduction of an MRL-kidney treatment.

5. Conclusion

The kidney position is more reproducible in expiration non-guided free breathing than in un-coached expiration breath hold for patients with a renal lesion. For a gated radiation treatmentof RCC, using a small gating window, free breathing seems to be the most time efficientbreathing pattern.

Acknowledgments

Part of this research is financially supported by the Dutch Technology Foundation STW.

References

Adler J R, Chang S D, Murphy M J, Doty J, Geis P and Hancock S L 1997 The Cyberknife: a frameless roboticsystem for radiosurgery Stereotact. Funct. Neurosurg. 69 124–8

Atwell T D et al 2012 Complications following 573 percutaneous renal radiofrequency and cryoablation proceduresJ. Vasc. Interv. Radiol. 23 48–54

Berson A M, Emery R, Rodriguez L, Richards G M, Ng T, Sanghavi S and Barsa J 2004 Clinical experience usingrespiratory gated radiation therapy: comparison of free-breathing and breath-hold techniques Int. J. Radiat.Oncol. Biol. Phys. 60 419–26

Bussels B, Goethals L, Feron M, Bielen D, Dymarkowski S, Suetens P and Haustermans K 2003 Respiration-inducedmovement of the upper abdominal organs: a pitfall for the three-dimensional conformal radiation treatment ofpancreatic cancer Radiother. Oncol. 68 69–74

Case R B, Sonke J J, Moseley D J, Kim J, Brock K K and Dawson L A 2009 Inter- and intrafraction variability inliver position in non-breath-hold stereotactic body radiotherapy Int. J. Radiat. Oncol. Biol. Phys. 75 302–8

Crijns S P M, Kok J G M, Lagendijk J J W and Raaymakers B W 2011 Towards MRI-guided linear acceleratorcontrol: gating on an MRI accelerator Phys. Med. Biol. 56 4815–25

2244 M K Stam et al

Crijns S P M, Raaymakers B W and Lagendijk J J W 2012 Proof of concept of MRI-guided tracked radiation delivery:tracking one-dimensional motion Phys. Med. Biol. 57 7863–72

Dawson L A, Brock K K, Kazanjian S, Fitch D, McGinn C J, Lawrence T S, Ten Haken R K and Balter J 2001 Thereproducibility of organ position using active breathing control (ABC) during liver radiotherapy Int. J. Radiat.Oncol. Biol. Phys. 51 1410–21

Dempsey J F, Benoit D, Fitzsimmons J R, Haghighat A, Li J G, Low D A, Mutic S, Palta J R, Romeijn H Eand Sjoden G E 2005 A device for realtime 3D image-guided IMRT Int. J. Radiat. Oncol. Biol. Phys.63 (Suppl. 1) S202

Dietrich L, Tcking T, Nill S and Oelfke U 2005 Compensation for respiratory motion by gated radiotherapy: anexperimental study Phys. Med. Biol. 50 2405–14

Eccles C, Brock K K, Bissonnette J P, Hawkins M and Dawson L A 2006 Reproducibility of liver position usingactive breathing coordinator for liver cancer radiotherapy Int. J. Radiat. Oncol. Biol. Phys. 64 751–9

Fallone B G, Murray B, Rathee S, Stanescu T, Steciw S, Vidakovic S, Blosser E and Tymofichuk D 2009 First MRimages obtained during megavoltage photon irradiation from a prototype integrated linac-MR system Med.Phys. 36 2084–8

Gill I S, Matin S F, Desai M M, Kaouk J H, Steinberg A, Mascha E, Thornton J, Sherief M H, Strzempkowski Band Novick A C 2003 Comparative analysis of laparoscopic versus open partial nephrectomy for renal tumorsin 200 patients J. Urol. 170 64–8

Jaffray D A, Siewerdsen J H, Wong J W and Martinez A A 2002 Flat-panel cone-beam computed tomography forimage-guided radiation therapy Int. J. Radiat. Oncol. Biol. Phys. 53 1337–49

Jhooti P, Haas T, Kawel N, Bremerich J, Keegan J and Scheffler K 2011 Use of respiratory biofeedback and clawsfor increased navigator efficiency for imaging the thoracic aorta Magn. Reson. Med. 66 1666–73

Keall P J, Kini V R, Vedam S S and Mohan R 2001 Motion adaptive x-ray therapy: a feasibility study Phys. Med.Biol. 46 1–10

Kerkhof E M, Raaymakers B W, van Vulpen M, Zonnenberg B A, Bosch J L H R, van Moorselaar R J Aand Lagendijk J J W 2011 A new concept for non-invasive renal tumour ablation using real-time MRI-guidedradiation therapy BJU Int. 107 63–8

Kini V R, Vedam S S, Keall P J, Patil S, Chen C and Mohan R 2003 Patient training in respiratory-gated radiotherapyMed. Dosim. 28 7–11

Klein S, Staring M, Murphy K, Viergever M A and Pluim J P W 2010 Elastix: a toolbox for intensity-based medicalimage registration IEEE Trans. Med. Imaging 29 196–205

Korreman S S, Pedersen A N, Nøttrup T J, Specht L and Nystrom H 2005 Breathing adapted radiotherapy for breastcancer: comparison of free breathing gating with the breath-hold technique Radiother. Oncol. 76 311–8

Kubo H D and Hill B C 1996 Respiration gated radiotherapy treatment: a technical study Phys. Med. Biol. 41 83–91Lagendijk J J W, Raaymakers B W, Raaijmakers A J E, Overweg J, Brown K J, Kerkhof E M, Van der Put R W,

Hardemark B, Van Vulpen M and Van der Heide U A 2008 MRI/linac integration Radiother. Oncol. 86 25–9Linthout N, Bral S, Van de Vondel I, Dirk V, Tournel K, Gevaert T, Duchateau M, Reynders T and Storme G

2009 Treatment delivery time optimization of respiratory gated radiation therapy by application of audio-visualfeedback Radiother. Oncol. 91 330–5

Moerland M A, Van den Bergh A C, Bhagwandien R, Janssen W M, Bakker C J, Lagendijk J J W and Battermann J J1994 The influence of respiration induced motion of the kidneys on the accuracy of radiotherapy treatmentplanning, a magnetic resonance imaging study Radiother. Oncol. 30 150–4

Nøttrup T J, Korreman S S, Pedersen A N, Aarup L R, Nystrom H, Olsen M and Specht L 2007 Intra- and interfractionbreathing variations during curative radiotherapy for lung cancer Radiother. Oncol. 84 40–8

Park S, Anderson J K, Matsumoto E D, Lotan Y, Josephs S and Cadeddu J A 2006 Radiofrequency ablation of renaltumors: intermediate-term results J. Endourol. 20 569–73

Peng Y, Vedam S, Chang J Y, Gao S, Sadagopan R, Bues M and Balter P 2011 Implementation of feedback-guidedvoluntary breath-hold gating for cone beam ct-based stereotactic body radiotherapy Int. J. Radiat. Oncol. Biol.Phys. 80 909–17

Raaymakers B W et al 2009 Integrating a 1.5 T MRI scanner with a 6 MV accelerator: proof of concept Phys. Med.Biol. 54 N229–37

Remouchamps V M, Letts N, Vicini F A, Sharpe M B, Kestin L L, Chen P Y, Martinez A A and Wong J W 2003Initial clinical experience with moderate deep-inspiration breath hold using an active breathing control devicein the treatment of patients with left-sided breast cancer using external beam radiation therapy Int. J. Radiat.Oncol. Biol. Phys. 56 704–15

Schwartz L H, Richaud J, Buffat L, Touboul E and Schlienger M 1994 Kidney mobility during respiration Radiother.Oncol. 32 84–6

Stam M K, Crijns S P M, Zonnenberg B A, Barendrecht M M, van Vulpen M, Lagendijk J J W and Raaymakers B W2012 Navigators for motion detection during real-time MRI-guided radiotherapy Phys. Med. Biol. 57 6797–805

Kidney motion during free breathing and breath hold for MR-guided radiotherapy 2245

Vedam S S, Keall P J, Kini V R and Mohan R 2001 Determining parameters for respiration-gated radiotherapy Med.Phys. 28 2139–46

Wong J W, Sharpe M B, Jaffray D A, Kini V R, Robertson J H, Stromberg J S and Martinez A A 1999 The use ofactive breathing control (ABC) to reduce margin for breathing motion Int. J. Radiat. Oncol. Biol. Phys. 44 911–9

Wyler S F, Sulser T, Ruszat R, Weltzien B, Forster T H, Provenzano M, Gasser T C and Bachmann A 2007Intermediate-term results of retroperitoneoscopy-assisted cryotherapy for small renal tumours using multipleultrathin cryoprobes Eur. Urol. 51 971–9