7/18/2014
Acknowledgments
Image-Guidance for SBRT:
Technical Challenges and Pitfalls
• Colleagues and SBRT Teams at Duke
• Research supports from NIH and Varian Medical
Systems
Fang-Fang Yin, PhD
Department of Radiation Oncology
Duke University Medical Center
AAPM July 2014
Imaging: 4D CT for Organ Motion
Outline:
Challenges:
• Poor soft-tissue contrast
• Only produce one breathing cycle
• Correlation between internal target and external surrogates
• Imaging dose …
• Imaging: through the process
• Margin: GTV  PTV
• Planning and delivery
• Treatment assessment and process management
Imaging: Motion with 4D CT and 4D MRI
4DCT
4D Imaging: Soring Types
•
•
•
•
4DMRI
Retrospective – sorting same phase or amplitude
Prospective – image at a specific phase or amplitude
Phase gated – sorting based on phase
Amplitude gated – sorting based on amplitude
Single slice cine MRI
How accurate is 4-D CT?
Cai et al
1
7/18/2014
4D Imaging: Retrospective Sorting
1st
table position
2nd
table position
3rd
4D Imaging: Related Images
table position
Phases
MIP
• FB/BH CT
•
10-phase FB 4DCT
•
Fluoroscopy
AIP
MinIP
Treatment
Planning
System
MIP
AIP
MinIP
CT-0%
CT-10%
phase
phase
CT-80% phase
Pan et al. Med Phys 2004
AIP (color)
Cover, et al, Red J, 2006
Imaging: Irregular Breathing CBCT
Imaging: 4DCT Sorting Artifact
ITV from irregular breathing patient
Tumor
Free-Breathing
ITV (cm3)
4D ITV
(cm3)
Volume
Underestimation
(%)
A
B
1.78
35.62
2.97
46.98
40.1
24.2
Small Tumor A
FB ITV
4D ITV
3.5mm
Errors from recorded breathing pattern from
surface marker – irregular pattern
Large Tumor B
FB ITV
4D ITV
I/E=0.31
Corrected
Deviated
2.5mm
Duke University
Vergalasova et al Med Phys 2011
Imaging: Position Effect on Fusion
Imaging: Volumes in Gated CT
Planning CT
Planning MRI
• Pictures show MIP images
• Pictures show 30-70% volume
• Pictures show free breathing volume
Free-breathing CT data
CT-MR Fusion
MR FOV limitations
Body constraints
40-50% PW CT data
Duke University
Duke University
3-4 cm
2
7/18/2014
Imaging: Treatment Planning
Imaging: CT for Dose Calculation?
Often, we have CTs for lung/liver SBRT:
• FB, 4DTC, AIP, MIP, and BH CT data sets
• Handling of Motion Images
– For contouring
– For dose calculation
– For localization (prior treat)
– For verification (during treat)
– For assessment (post treat)
Freebreathing
Lung - Pretreatment Localization
Duke University
CBCT prior
correction
Average intensity
Maximum intensity
projection
projection
Yuan Tian et al, Med Phys 2012
Imaging: Overall Assessment
• A spinal SRS/SBRT was imaged three times
ITVCBCT
– Prior to correction
– After correction before treat
– After treatment
A
B
C
• Factors determining the margin
–
–
–
–
–
Planning CT
PTV
Setup uncertainty
Organ motion uncertainty
CTV motion trajectories
IGRT system errors
RTOG0631
CBCT after
correction
Imaging: Gated Treatment Volumes
MIP_GatedPhases
4DCT
MIP_AllPhases (0%-90%)
MIP_GatedPhases (30%-70%)
AIP_AllPhase (0%-90%)
AIP_GatedPhases (30%-70%)
Imaging: Treatment Verification
MIP_AllPhases
PTV
CBCT-ITV for online match
Dose calculation
Duke University
Gating window
Verification of
implanted marker
Duke University
3
7/18/2014
Imaging: Real-time Treat Verification
Margin: Setup Uncertainty
Goal for immobilization
• Minimize patient motion
Alpha cradle, arm-up
Wing board, arm-up
• Minimize organ motion Alpha cradle, arm-up
• Comfortable, stable, reproducible, or
predictable motion
Implanted
marker
Anatomical
marker
1.5 cm
0.7 cm
0.3 cm
Duke University
Margin: Organ Motion Uncertainty
Active breath control:
The residual errors of GTV
ML: 0.3±1.8 mm
AP: 1.2±2.3 mm
SI: 1.1±3.5 mm
Cheung, et al, Red J 2003
• Remains some inter–breath hold variability in
peripheral lung
• Limited reduction of PTV margin
Margin: Organ Motion Uncertainty
Total intravenous
anesthesia (TIVA)
High-frequency jet
ventilation (HFJV)
Animal study:
Motion range: < 3 mm
Duke University
Margin: Motion Uncertainty
Abdominal Compression
Mean motion reduction:
3.5 mm for lower lobe tumors
0.8 mm for upper/middle lobe
Sometime, compression increased
tumor motion
Mean ITV reduction:
3.6 cc for lower lobe lesions
0.2 cc for upper/middle lobe lesions
Dosimetric gain for lung sparing was not clinically relevant
Bouilhol et al, 2012, Phys Med
Margin: Volume Uncertainty
ITV10phase (blue line)
ITVMIP (green line)
GTV3D (red line)
Need multiple
CT data
Yin et al 2001 IJROBP:
“Extracranial radiosurgery:
Immobilizing liver motion using
high-frequency jet ventilation and
total intravenous anesthesia”
Ge et al, Red J 2012
4
7/18/2014
Margin: Imaging System Uncertainty
Margin: Breathold Uncertainty
Breath hold 1
3 mm
rotation
3 mm
shift
Breath hold 2
Breath hold 3
Cord dose
>20%
Duke University
Delivery: Smart Choice of VMAT/IMRT
Margin needed ~ 5 mm
Duke University
Delivery: MLC Leaf Width
5 mm
Relative Volume (%)
5 mm
RapidArc MU = 714
IMRT Total MU = 1572
2.5 mm
0
Time: 1.3 min
Duke University
Delivery: MLC Interplay Effect
2.5 mm
10
20
30
40
50
60
70
80
90 100 110
Relative dose (%)
Wu et al Radiat Oncol 2009
Delivery: High Doserate (FFF)
Beam-on timing
Doserate ~2400 MU/min
Beam-on @ different points
in breathing period
Differences in
delivered dose
S Benedict AAPM 2012
Any biological effect?
5
7/18/2014
Assessment: Adaptive Volume Changes
Process Management for SBRT
Before
After
Wu et al PMB 2008
Salama, Kirkpatrick, and Yin
Nature Reviews|Clinical Oncology 2012
Thank you for your attention
6