|
Medical Imaging |
Cancer is a disease in which cells divide uncontrollably, invading local and distant (metastases) structures. This lack of normal function of cells is predominantly caused by genetic damage.
Within the UK more than 250,000 new cases are diagnosed yearly. For the UK population it is considered that more than 1 in 3 will develop cancer within their lifetime [1].
There are three main treatments available to cancer patients, and selection of the most appropriate single or combination of treatment is dependant upon several factors such as cancer site, cancer type, stage and patient health.
• Surgery - the oldest form of cancer treatment. It
is a highly invasive procedure to physically excise the diseased tissue.
• Chemotherapy – the targeted attack of cancer cells using chemicals.
Administered intravenously or orally, chemotherapy drugs attack cancer cells
during mitosis by damaging essential proteins or DNA. Since cancer cells divide
much more rapidly than the majority of normal cells, chemotherapeutic drugs
are largely successful. Undesirable side effects such as hair and nausea occur
because certain normal cells which divide rapidly, such as hair follicles and
bone marrow, are affected by the drugs.
• Radiotherapy – the use of ionising radiation to damage cell DNA
and prevent cell division. External beam radiotherapy is delivered by a linear
accelerator over a course of fractions. In brachytherapy, radioactive sources
are surgically implanted into the tumour.
Alternative cancer treatments include gene therapy, immunotherapy and hormone therapy.
Patient imaging is an important first step of planning treatment. Treatment strategies are largely based on information gained from patient scans. The main imaging modalities used are:
• X-ray – high energy electromagnetic radiation
is directed at the site to be investigated. This beam is attenuated by differing
mass and this mass attenuation is detected on a photographic plate.
• Computed Tomography (CT) – the patient lies in a CT scanner and
multiple x-rays are taken in a 360º arc around the region of interest.
The mass attenuation is measured continuously and used to reconstruct an axial
slice. Multiple slices are generally taken at variable slice separation. The
use of kV x-rays allows good image quality and soft tissue contrast.
• Magnetic Resonance (MR) – the measurement of proton spins in water
molecules by using large magnetic fields. The patient is placed in a large magnetic
field and then pulses of radiowaves are given to the subject, which are in turn
re-emitted containing information about the make-up of the internal structure.
• Positron Emission Tomography (PET) – small amounts of radiolabelled
drugs are administered intravenously to the patient which accumulate in the
tumour. As the radioactive element decays, the emitted positron annihilates
with an electron to give two gamma rays at 180º which are detected and
used to locate the tumour.
• Ultrasound – low frequency sound waves are used to generate an
internal picture of the patient. The waves are generated and recorded by a scanning
head placed in direct contact with the patient’s skin which measures waves
reflected off internal structures to build up an echo picture.
Other imaging modalities include bone scans and Single Photon Emission Computed Tomography (SPECT).
a. Computed Tomography (CT)
The first commercially available CT scanners appeared in 1972 and came from
collaboration between the EMI Company, Atkinson Morley's Hospital, London and
the Department of Health and Social Security (DHSS) [2]. Geoffrey Hounsfield
conceived and developed the idea in the late 1960s that "a realisation
that you could determine what was in a box by taking readings at all angles
through it". Allan Cormack, a South African physicist was also working
on a CT scanner, but not in collaboration with Hounsfield. The early clinical
CT scanners took several minutes to acquire data and much longer to compute
the complicated mathematics needed to reconstruct images, with a low resolution.
With the advances in this technology, modern CT scanners can perform multi-slice
CT scans and reconstruct high resolution images in seconds.
b. Spiral CT
In conventional CT a tomography is acquired by rotating the scanner through
one revolution about a fixed location on the patient’s body. In spiral
CT image acquisition, the gantry performs several rotations whilst gradually
moving along the axis of the patient. The resulting image is that of a volume
of the patient.
c. Multi-slice CT
In 1992 the first clinical multi-slice scanner became available. As the name
suggests, a multi-slice CT scanner acquires multiple slices through one revolution
using multiple arrays of parallel detectors. Compared with conventional CT scanners,
multi-slice CT scanners have the advantage of increased spatial resolution along
the axis of the patient and decreased scanning time.
d. Cone Beam CT
Conventional CT scanners use a thin width x-ray beam to image the patient. This
beam is created from a point source kV x-ray generator which is then collimated,
and the detector is usually of the order of millimetres.
In cone beam CT scanning, the angle of the imaging beam (collimator) is increased so that the beam forms a cone. The axial length of this beam is also increased, and the detector used is a flat panel. The advantage given by the cone beam CT image is that the recorded data can be reconstructed to give a 3D volumetric image, compared with the single slice of the conventional CT scan.
Figure 1. Computed Tomography scan versus Cone Beam Computed Tomography.
Figure 2. Geometry correction for cone beam.
This extra information does come at a cost however. The algorithms needed to compensate for geometric distortions due to the nature of the beam mean that greater computing power and reconstruction time is needed.
Much of the development work of this clinical cone beam imaging was carried out by Jaffray, Wong et al. at William Beaumont Hospital, Royal Oak, MI, USA. The group described their initial experience with a cone beam imager attached to a linear accelerator [3, 4]. They describe how a flat panel kV image detector was attached to a standard linear accelerator, assembled 90º to the MV treatment beam. Both the kV x-ray source and the detector were retractable so as not to cause any interference with the normal operation of the treatment machine.
Advantages gained from incorporating the scanner onto the treatment machine include the patient can be imaged in the treatment position and the image is automatically in the treatment beam’s frame of reference.
In oncology the CT scan is a common radiographic tool used by clinicians as a method to gain information regarding the exact position and volume of a tumour, and its relation to other structures and Organs at Risk (OARs). Generally, each cancer centre will have protocols for specific cancer types that describe dose, fractionation and the number of fields to use, but how the dose is delivered is specific to the patient’s physiology, which is known to millimetre accuracy through CT scans.
Using specialised software, radiotherapy treatment is planned on multiple CT slices from the RTP scan, on which Gross Tumour Volumes (GTV), Clinical Target Volumes (CTV) and Planning Target Volumes (PTV) are outlined.
The GTV is the exact volume of the tumour, and this boundary is delineated on the scan. Where the tumour cannot be outlined, e.g. after surgical removal, the clinician outlines the CTV which is the volume which encompasses the GTV, when present, plus any area where microscopic disease and spread are likely to be present. The PTV is the CTV plus a deliberate margin to include normal tissue. This margin is to allow for organ motion and deformation and treatment set-up errors. Margin size is dependent upon organ involved and international guidelines are available [5].
Treatment is planned over the necessary number of CT slices as the target region covers, with an average slice separation of 5mm.
Figure 3. CT Slices through a structure
The PTV is the volume which must contain 95% of the planned dose, and the CTV 99%. The field angles used are based upon the relative position of OARs and maximum delivered doses to these are a major factor.
The RTP scan is a single snapshot of the patient in the treatment position, generally one to two weeks before treatment begins. Curative treatment courses are generally several weeks in length.
Technical advances in delivery of radiotherapy over recent years mean that highly conformal beams can now be generated to deliver curative doses [many]. Highly conforming beams mean that less normal tissue is irradiated and higher doses can be delivered safely to the target volume. Studies show that the higher dose to tumour increases tumour control probability [need ref].
The Multi-Leaf Collimator (MLC) has dramatically increased precision of treatment beams and fields. Prior to the MLC, crude rectangular fields were only possible which delivered dose to large regions of normal tissue. Some shielding in the form of lead blocks could be used to protect vital organs (e.g. heart) form the fields. The MLC consists of a head with up to 120 individual leaves that can be inserted precise distances into the treatment field, thus creating precise field shapes.
a. Organ Motion and Deformation
Organ motion and deformation can broadly be split into two categories: inter-fraction;
and intra-fraction.
Inter-fraction motion and deformation is that which occurs between individual
treatment fractions includes filling (e.g. bladder, rectal), inflammation and
atrophy.
Intra-fraction motion and deformation is that which occurs within a single treatment
fraction and is mainly due to filling, peristalsis and breathing.
To combat the effects of bladder filling drinking protocols can be used. These are instructions given to the patient detailing how much liquid to consume at a certain time before the planning scan and each subsequent treatment fraction, for treatment based on a full bladder. An alternative to this is for the patient to empty their bladder before the RTP scan and each fraction.
To reduce breathing motion and deformation one method is to gate treatment and use the “step and shoot” technique whereby the treatment beam is only on in certain phases of the breathing cycle. This increases patient treatment time and is reliant upon a consistent, reproducible breath cycle. Another method which controls breathing motion is the commercially available Active Breathing Coordinator (ABC) device (Elekta Oncology Systems, Crawley, Surrey). The patient’s breathing is controlled by the device and is stopped at the same instance in either the inspiration or expiration cycle and the treatment is then delivered.
b. Set-up Reproducibility
Wall and ceiling mounted lasers are commonly used in radiotherapy treatment
rooms as means of setting-up the patient in the correct treatment position.
For certain cancer sites (e.g. breast, lung, prostate) a small tattoo is placed
on the skin where the laser is incident upon the patient. These tattoos are
used as markers in subsequent treatment fractions by the radiographers to position
the patient correctly. These tattoos are used as surrogates for internal organ
and tumour positions.
Electronic Portal Imaging Devices (EPIDs) are MV cameras that are used as another means to verify that the patient is in the correct treatment position. The image gained is from the treatment beam’s eye view and so gives accurate spatial information. However the power of the beam means that soft tissue cannot be viewed but more dense structures i.e. bones are visible. The bony landmarks on the EPI image can be compared to those on the RTP scan and this information used to move the patient if its correlation to that of the RTP scan is sufficiently misaligned.
Similar to bony landmark matching, another technique used to verify patient set-up is imaging of implanted radio-opaque markers. Some cancers (e.g. breast, pancreas, prostate) can have gold seeds implanted [6] or surgical clips inserted prior to radiotherapy, which can be imaged and used to verify set-up.
Each of these verification techniques use surrogates for tumour position. Errors in these techniques can occur because skin can move independently of internal organs e.g. weight loss, bones can move, e.g. femoral head rotation, and implanted markers can migrate.
One of the most important recent advances in radiotherapy has been the development of Intensity Modulated Radiotherapy (IMRT). Inverse treatment planning allows the clinician to outline the CTV and PTV on the RTP scan and set the dose required, number of treatment beams and maximum dose permissible to sensitive organs. The computer then generates several possible treatment plans to achieve the requirements. The computer controls the MLC to tightly conform the field shape. This allows for the treatment margins to be minimised and therefore spare maximum normal tissue.
Studies have shown that local tumour control rates are increased if the dose to the tumour is increased during the course radiotherapy [7, 8].
The importance of localisation of the target volume for radiotherapy was well understood before the clinical availability of cone beam CT. In 2001 Martinez et al. described how they developed a method using EPIDs and CT images to create a patient specific PTV which can be used to safely dose escalate and improve the confidence with the use of IMRT for locally advanced prostate cancer patients [9]. They describe an Adaptive Radiotherapy (ARP) approach whereby the first step (used in this study) target/treatment image feedback is used to correct for random and systematic set-up variations to the treatment plan. Daily EPI and CT images of the prostate were taken and an off-line correction made at the end of week one. Correlation of the second plan compared with the original was used to determine the method of treatment; IMRT for closely correlating plans, and Conformal Radiotherapy (CRT) for larger differences. One hundred and fifty patients were treated using the ART process. The amount of dose delivered using either CRT or IMRT compared to that which would have been delivered if generic margins (PTV) had been used were calculated. For CRT an increase of 5% dose was given to the target volume and for IMRT an average of 7.5% extra dose was given. This paper shows that when daily images of the target volume are taken and used to re-plan the patient after five days of treatment, the benefit in terms of dose escalation can be significant.
A different approach to account for prostate movement is
described by Wong et al. [10]. The group used an in treatment room CT scanner
to image patients prior to treatment. This image was then compared to the RTP
scan and used to derive a new isocentre, and treat the patient.
One hundred and eight patients were treated using this method and 540 daily
CT scans performed. Patients were only treated during their final week of radiotherapy
as severe or chronic damage to the rectal wall would only occur at this point
in treatment, should it have been included in the high dose volume. The corrections
made due to prostate position on day of scan were recorded and are shown in
table 1.
| Movement of Prostate (mm) | AP Direction (%) | SI Direction (%) | Left-Right Direction (%) |
| No Adjustment | 46 | 73 | 66 |
| 3-4 | 10 | 2 | 10 |
| 5-9 | 29 | 21 | 19 |
| >= 10 | 15 | 4 | 5 |
Table 1. Percent adjustment of the isocentre as obtained by daily CT scans.
Abbreviations: AP = Anterior - Posterior; SI = Superior - Inferior
Table taken from [10].
This data shows the range of prostate movement and the importance of quantifying and correcting for it in the prevention of morbidity of normal tissue.
Smithmans et al. used repeat CT images of prostate cancer patients to develop
an automatic localisation method for the organ [11]. Nineteen patients were
given between 8 – 13 repeat CT scans during treatment. The RTP scan plus
repeat CT scan were co-registered using bony landmarks. Next the prostate outline
from the RTP scan was given a 5mm margin to define a region of interest large
enough to give grey level contrast information necessary for registration. This
region was then automatically registered to a repeat CT scan and 3D grey value
registration used to localise the prostate. This automatically registered prostate
region was then compared to manually outlined versions. The overall success
rate of this automatic registration was 91% and the registration was performed
in 45 seconds on a standard desktop computer.
This information is very important in terms of the development of on-line correction
strategies. It is the first study to show that automatic localisation of target
organs can be performed in a timescale that would make an on-line correction
protocol feasible.
Ghilezan et al. describe a study in which they used multiple
serial CT slices to approximate to cone beam images for the theoretical benefit
gained from using Image Guided Intensity Modulated Radiotherapy (IG-IMRT) for
prostate cancer patients [12]. Twenty two patients undergoing conventional radiotherapy
for prostate cancer were scanned on multiple occasions (median 18 per patient).
An IMRT plan was made for each patient from their RTP scan and an IG-IMRT plan
created using the scan of the day. The equivalent uniform dose to the target
volume for the IMRT and IG-IMRT plans were compared and the average dose increment
permissible by the on-line IG-IMRT technique compared with the IMRT was 13%.
This study quantifies the potential benefit of IG-IMRT over IMRT, however a
limiting factor in the validity of the results is that intra-fraction motion
was not taken into consideration. This ideal value of 13% increase in dose cannot
be translated reliably into the real world where intra-fraction motion plays
a major part in radiotherapy treatment and tumour control.
Deurloo et al. describe their study to measure the position of the prostate and seminal vesicles during a course of radiotherapy treatment using the RTP scan and repeat CT scans of nineteen prostate cancer patients [13]. The patients received 8 – 12 repeat CT scans (average 11) and a single observer outlined the prostate and seminal vesicles on all scans, which eliminated intra-observer variability. These data were then to compute GTVs for each patient and a general group GTV. Shape variations between these two volumes were computed and the conclusion drawn was that during the course of radiotherapy treatment for prostate cancer, the variations in prostate and seminal vesicle position due to organ deformation are small when compared to patient set-up errors and organ motion.
A search on PubMed [14] with the search terms “image guided radiotherapy” returns 263 results. With the search terms “cone beam CT” returns 177 results. A search with the terms “image guided radiotherapy AND cone beam CT” returns 5 results, all of which are from the last three years.
With both IMRT and dose escalation protocols being used regularly, it is now more important than ever that the localisation of the CTV is verified. If the PTV is decreased to spare normal tissue, it is essential to verify that the CTV is still encompassed within the PTV, and that this reduction is not at the expenses of tumour control probability. The use of cone beam CT for on-line image guided radiotherapy is currently in its preliminarily research stage, but the potential it holds for the future is tremendous.
1. Cancer Research UK Information Resource
Centre (2005), CancerStats, http//:info.cancerresearchuk.org/cancerstats
2. The British Journal of Radiology, Vol 65, Issue 777 835-837
3. Jaffray, D. A., D. G. Drake, et al. (1999). "A radiographic
and tomographic imaging system integrated into a medical linear accelerator
for localization of bone and soft-tissue targets." Int J Radiat Oncol Biol
Phys 45(3): 773-89.
4. Jaffray, D. A., J. H. Siewerdsen, et al. (2002). "Flat-panel
cone-beam computed tomography for image-guided radiation therapy." Int
J Radiat Oncol Biol Phys 53(5): 1337-49.
5. ICRU Report 50. Dose Specification for Reporting External
Beam Therapy with Photons and Electrons. International Commission on Radiation
Units and Measurements. Washington DC, 1978.
6. Henry AM, Wilkinson C, Wylie JP, Logue JP, Price P, Khoo
VS. Trans-perineal implantation of radio-opaque treatment verification markers
into the prostate: an assessment of procedure related morbidity, patient acceptability
and accuracy. Radiother Oncol. 2004 Oct;73(1):57-9.
7. Dearnaley, D. P., E. Hall, et al. (2005). "Phase III
pilot study of dose escalation using conformal radiotherapy in prostate cancer:
PSA control and side effects." Br J Cancer 92(3): 488-98.
8. Pollack, A., G. K. Zagars, et al. (2002). "Prostate
cancer radiation dose response: results of the M. D. Anderson phase III randomized
trial." Int J Radiat Oncol Biol Phys 53(5): 1097-105.
9. Martinez, A. A., D. Yan, et al. (2001). "Improvement
in dose escalation using the process of adaptive radiotherapy combined with
three-dimensional conformal or intensity-modulated beams for prostate cancer."
Int J Radiat Oncol Biol Phys 50(5): 1226-34.
10. Wong, J. R., L. Grimm, et al. (2005). "Image-guided
radiotherapy for prostate cancer by CT-linear accelerator combination: prostate
movements and dosimetric considerations." Int J Radiat Oncol Biol Phys
61(2): 561-9.
11. Smitsmans, M. H., J. W. Wolthaus, et al. (2004). "Automatic
localization of the prostate for on-line or off-line image-guided radiotherapy."
Int J Radiat Oncol Biol Phys 60(2): 623-35.
12. Ghilezan, M., D. Yan, et al. (2004). "Online image-guided
intensity-modulated radiotherapy for prostate cancer: How much improvement can
we expect? A theoretical assessment of clinical benefits and potential dose
escalation by improving precision and accuracy of radiation delivery."
Int J Radiat Oncol Biol Phys 60(5): 1602-10.
13. Deurloo, K. E., R. J. Steenbakkers, et al. (2005). "Quantification
of shape variation of prostate and seminal vesicles during external beam radiotherapy."
Int J Radiat Oncol Biol Phys 61(1): 228-38.
14. PubMed, a service of the National Library of Medicine,
includes over 15 million citations for biomedical articles back to the 1950's.
http://www.ncbi.nlm.nih.gov/Database/index.html
| ©2005 Jimoid.com |