CTV-to-PTV margin assessment for esophageal cancer radiotherapy based on an accumulated dose analysis.

PURPOSE: This study aimed to assess the smallest clinical target volume (CTV) to planned target volume (PTV) margins for esophageal cancer radiotherapy using daily online registration to the bony anatomy that yield full dosimetric coverage over the course of treatment. METHODS: 29 esophageal cancer patients underwent six T2-weighted MRI scans at weekly intervals. An online bone-match image-guided radiotherapy treatment of five fractions was simulated for each patient. Multiple conformal treatment plans with increasing margins around the CTV were created for each patient. Then, the dose was warped to obtain an accumulated dose per simulated fraction. Full target coverage by 95% of the prescribed dose was assessed as a function of margin expansion in six directions. If target coverage in a single direction was accomplished, then the respective margin remained fixed for the subsequent dose plans. Margins in uncovered directions were increased in a new dose plan until full target coverage was achieved. RESULTS: The smallest set of CTV-to-PTV margins that yielded full dosimetric CTV coverage was 8 mm in posterior and right direction, 9 mm in anterior and cranial direction and 10 mm in left and caudal direction for 27 out of 29 patients. In two patients the curvature of the esophagus considerably changed between fractions, which required a 17 and 23 mm margin in right direction. CONCLUSION: Accumulated dose analysis revealed that CTV-to-PTV treatment margins of 8, 9 and 10 mm in posterior & right, anterior & cranial and left & caudal direction, respectively, are sufficient to account for interfraction tumor variations over the course of treatment when applying a daily online bone match. However, two patients with extreme esophageal interfraction motion were insufficiently covered with these margins and were identified as patients requiring replanning to achieve full target coverage.

Gross tumour delineation on computed tomography and positron emission tomography-computed tomography in oesophageal cancer: A nationwide study.

BACKGROUND AND PURPOSE: Accurate delineation of the primary tumour is vital to the success of radiotherapy and even more important for successful boost strategies, aiming for improved local control in oesophageal cancer patients. Therefore, the aim was to assess delineation variability of the gross tumour volume (GTV) between CT and combined PET-CT in oesophageal cancer patients in a multi-institutional study. MATERIALS AND METHODS: Twenty observers from 14 institutes delineated the primary tumour of 6 cases on CT and PET-CT fusion. The delineated volumes, generalized conformity index (CIgen) and standard deviation (SD) in position of the most cranial/caudal slice over the observers were evaluated. For the central delineated region, perpendicular distance between median surface GTV and each individual GTV was evaluated as in-slice SD. RESULTS: After addition of PET, mean GTVs were significantly smaller in 3 cases and larger in 1 case. No difference in CIgen was observed (average 0.67 on CT, 0.69 on PET-CT). On CT cranial-caudal delineation variation ranged between 0.2 and 1.5 cm SD versus 0.2 and 1.3 cm SD on PET-CT. After addition of PET, the cranial and caudal variation was significantly reduced in 1 and 2 cases, respectively. The in-slice SD was on average 0.16 cm in both phases. CONCLUSION: In some cases considerable GTV delineation variability was observed at the cranial-caudal border. PET significantly influenced the delineated volume in four out of six cases, however its impact on observer variation was limited.

A novel digital workflow to manufacture personalized three-dimensional-printed hollow surgical obturators after maxillectomy.

Partial or complete resection of the maxilla during tumour surgery causes oronasal defects, leading to oral-maxillofacial dysfunction, for which the surgical obturator (SO) is an important treatment option. Traditional manufacturing of SOs is complex, time-consuming, and often results in inadequate fit and function. This technical note describes a novel digital workflow to design and manufacture a three-dimensional (3D)-printed hollow SO. Registered computed tomography and magnetic resonance imaging images are used for gross tumour delineation. The produced RTStruct set is exported as a stereolitography (STL) file and merged with a 3D model of the dental status. Based on these merged files, a personalized and hollow digital SO design is created, and 3D printed. Due to the proper fit of the prefabricated SO, a soft silicone lining material can be used during surgery to adapt the prosthesis to the oronasal defect, instead of putty materials that are not suitable for this purpose. An STL file of this final SO is created during surgery, based on a scan of the relined SO. The digital workflow results in a SO weight reduction, an increased fit, an up-to-date digital SO copy, and overall easier clinical handling.

Inter-observer agreement in GTV delineation of bone metastases on CT and impact of MR imaging: A multicenter study.

BACKGROUND AND PURPOSE: The use of Stereotactic Body Radiotherapy (SBRT) for bone metastases is increasing rapidly. Therefore, knowledge of the inter-observer differences in tumor volume delineation is essential to guarantee precise dose delivery. The aim of this study is to compare inter-observer agreement in bone metastases delineated on different imaging modalities. MATERIAL AND METHODS: Twenty consecutive patients with bone metastases treated with SBRT were selected. All patients received CT and MR imaging in treatment position prior to SBRT. Five observers from three institutions independently delineated gross tumor volume (GTV) on CT alone, CT with co-registered MRI and MRI alone. Four contours per imaging modality per patient were available, as one set of contours was shared by 2 observers. Inter-observer agreement, expressed in generalized conformity index [CIgen], volumes of contours and contours center of mass (COM) were calculated per patient and imaging modality. RESULTS: Mean GTV delineated on MR (45.9±52.0cm3) was significantly larger compared to CT-MR (40.2±49.4cm3) and CT (34.8±41.8cm3). A considerable variation in CIgen was found on CT (mean 0.46, range 0.15-0.75) and CT-MRI (mean 0.54, range 0.17-0.71). The highest agreement was found on MRI (mean 0.56, range 0.20-0.77). The largest variations of COM were found in anterior-posterior direction for all imaging modalities. CONCLUSIONS: Large inter-observer variation in GTV delineation exists for CT, CT-MRI and MRI. MRI-based GTV delineation resulted in larger volumes and highest consistency between observers.

First patients treated with a 1.5 T MRI-Linac: clinical proof of concept of a high-precision, high-field MRI guided radiotherapy treatment.

The integration of 1.5 T MRI functionality with a radiotherapy linear accelerator (linac) has been pursued since 1999 by the UMC Utrecht in close collaboration with Elekta and Philips. The idea behind this integrated device is to offer unrivalled, online and real-time, soft-tissue visualization of the tumour and the surroundings for more precise radiation delivery. The proof of concept of this device was given in 2009 by demonstrating simultaneous irradiation and MR imaging on phantoms, since then the device has been further developed and commercialized by Elekta. The aim of this work is to demonstrate the clinical feasibility of online, high-precision, high-field MRI guidance of radiotherapy using the first clinical prototype MRI-Linac. Four patients with lumbar spine bone metastases were treated with a 3 or 5 beam step-and-shoot IMRT plan. The IMRT plan was created while the patient was on the treatment table and based on the online 1.5 T MR images; pre-treatment CT was deformably registered to the online MRI to obtain Hounsfield values. Bone metastases were chosen as the first site as these tumors can be clearly visualized on MRI and the surrounding spine bone can be detected on the integrated portal imager. This way the portal images served as an independent verification of the MRI based guidance to quantify the geometric precision of radiation delivery. Dosimetric accuracy was assessed post-treatment from phantom measurements with an ionization chamber and film. Absolute doses were found to be highly accurate, with deviations ranging from 0.0% to 1.7% in the isocenter. The geometrical, MRI based targeting as confirmed using portal images was better than 0.5 mm, ranging from 0.2 mm to 0.4 mm. In conclusion, high precision, high-field, 1.5 T MRI guided radiotherapy is clinically feasible.

Planning, optimisation and evaluation of hyperthermia treatments.

BACKGROUND: Hyperthermia treatment planning using dedicated simulations of power and temperature distributions is very useful to assist in hyperthermia applications. This paper describes an advanced treatment planning software package for a wide variety of applications. METHODS: The in-house developed C++ software package Plan2Heat runs on a Linux operating system. Modules are available to perform electric field and temperature calculations for many heating techniques. The package also contains optimisation routines, post-treatment evaluation tools and a sophisticated thermal model enabling to account for 3D vasculature based on an angiogram or generated artificially using a vessel generation algorithm. The use of the software is illustrated by a simulation of a locoregional hyperthermia treatment for a pancreatic cancer patient and a spherical tumour model heated by interstitial hyperthermia, with detailed 3D vasculature included. RESULTS: The module-based set-up makes the software flexible and easy to use. The first example demonstrates that treatment planning can help to focus the heating to the tumour. After optimisation, the simulated absorbed power in the tumour increased with 50%. The second example demonstrates the impact of accurately modelling discrete vasculature. Blood at body core temperature entering the heated volume causes relatively cold tracks in the heated volume, where the temperature remains below 40 °C. CONCLUSIONS: A flexible software package for hyperthermia treatment planning has been developed, which can be very useful in many hyperthermia applications. The object-oriented structure of the source code allows relatively easy extension of the software package with additional tools when necessary for future applications.

Registration of structurally dissimilar images in MRI-based brachytherapy.

A serious challenge in image registration is the accurate alignment of two images in which a certain structure is present in only one of the two. Such topological changes are problematic for conventional non-rigid registration algorithms. We propose to incorporate in a conventional free-form registration framework a geometrical penalty term that minimizes the volume of the missing structure in one image. We demonstrate our method on cervical MR images for brachytherapy. The intrapatient registration problem involves one image in which a therapy applicator is present and one in which it is not. By including the penalty term, a substantial improvement in the surface distance to the gold standard anatomical position and the residual volume of the applicator void are obtained. Registration of neighboring structures, i.e. the rectum and the bladder is generally improved as well, albeit to a lesser degree.

Expert-driven label fusion in multi-atlas-based segmentation of the prostate using weighted atlases.

PURPOSE: Automated segmentation is required for radiotherapy treatment planning, and multi-atlas methods are frequently used for this purpose. The combination of multiple intermediate results from multi-atlas segmentation into a single segmentation map can be achieved by label fusion. A method that includes expert knowledge in the label fusion phase of multi-atlas-based segmentation was developed. The method was tested by application to prostate segmentation, and the accuracy was compared to standard techniques. METHODS: The selective and iterative method for performance level estimation (SIMPLE) algorithm for label fusion was modified with a weight map given by an expert that indicates the importance of each region in the evaluation of segmentation results. Voxel-based weights specified by an expert when performing the label fusion step in atlas-based segmentation were introduced into the modified SIMPLE algorithm. These weights incorporate expert knowledge on accuracy requirements in different regions of a segmentation. Using this knowledge, segmentation accuracy in regions known to be important can be improved by sacrificing segmentation accuracy in less important regions. Contextual information such as the presence of vulnerable tissue is then used in the segmentation process. This method using weight maps to fine-tune the result of multi-atlas-based segmentation was tested using a set of 146 atlas images consisting of an MR image of the lower abdomen and a prostate segmentation. Each image served as a target in a set of leave-one-out experiments. These experiments were repeated for a weight map derived from the clinical practice in our hospital. RESULTS: The segmentation accuracy increased 6 % in regions that border vulnerable tissue using expert-specified voxel-based weight maps. This was achieved at the cost of a 4 % decrease in accuracy in less clinically relevant regions. CONCLUSION: The inclusion of expert knowledge in a multi-atlas-based segmentation procedure was shown to be feasible for prostate segmentation. This method allows an expert to ensure that automatic segmentation is most accurate in critical regions. This improved local accuracy can increase the practical value of automatic segmentation.

Patient position verification using small IMRT fields.

A commonly used approach to quantify and minimize patient setup errors is by using electronic portal imaging devices (EPIDs). The position of the tumor can be verified indirectly by matching the bony anatomy to a reference image containing the same structures. In this paper we present two off-line methods for detecting the position of the bony anatomy automatically, even if every single portal image of each segment of an IMRT treatment beam contains insufficient matching information. Extra position verification fields will no longer be necessary, which reduces the total dose to the patient. The first method, the stack matching method (SMM), stacks the portal image of each segment of a beam to a three dimensional (3D) volume, and this volume is subsequently used during the matching phase. The second method [the averaged projection matching method (APMM)], is a simplification of the first one, since the initially created volume is reduced again to a 2D artificial image, which speeds up the matching procedure considerably, without a significant loss of accuracy. Matching is based on normalized mutual information. We demonstrate our methods by comparing them to existing matching routines, such as matching based on the largest segment. Both phantom and patient experiments show that our methods are comparable with the results obtained from standard position verification methods. The matches are verified by means of visual inspection. Furthermore, we show that when a distinct area of 40-60 cm2 of the EPID is exposed during one treatment beam, both SMM and APMM are able to deliver a good matching result.

Assessment of rigid multi-modality image registration consistency using the multiple sub-volume registration (MSR) method.

Registration of different imaging modalities such as CT, MRI, functional MRI (fMRI), positron (PET) and single photon (SPECT) emission tomography is used in many clinical applications. Determining the quality of any automatic registration procedure has been a challenging part because no gold standard is available to evaluate the registration. In this note we present a method, called the 'multiple sub-volume registration' (MSR) method, for assessing the consistency of a rigid registration. This is done by registering sub-images of one data set on the other data set, performing a crude non-rigid registration. By analysing the deviations (local deformations) of the sub-volume registrations from the full registration we get a measure of the consistency of the rigid registration. Registration of 15 data sets which include CT, MR and PET images for brain, head and neck, cervix, prostate and lung was performed utilizing a rigid body registration with normalized mutual information as the similarity measure. The resulting registrations were classified as good or bad by visual inspection. The resulting registrations were also classified using our MSR method. The results of our MSR method agree with the classification obtained from visual inspection for all cases (p < 0.02 based on ANOVA of the good and bad groups). The proposed method is independent of the registration algorithm and similarity measure. It can be used for multi-modality image data sets and different anatomic sites of the patient.

Integrating a MRI scanner with a 6 MV radiotherapy accelerator: dose deposition in a transverse magnetic field.

Integrating magnetic resonance imaging (MRI) functionality with a radiotherapy accelerator can facilitate on-line, soft-tissue based, position verification. A technical feasibility study, in collaboration with Elekta Oncology Systems and Philips Medical Systems, led to the preliminary design specifications of a MRI accelerator. Basically the design is a 6 MV accelerator rotating around a 1.5 T MRI system. Several technical issues and the clinical rational are currently under investigation. The aim of this paper is to determine the impact of the transverse 1.5 T magnetic field on the dose deposition. Monte Carlo simulations were used to calculate the dose deposition kernel in the presence of 1.5 T. This kernel in turn was used to determine the dose deposition for larger fields. Also simulations and measurements were done in the presence of 1.1 T. The pencil beam dose deposition is asymmetric. For larger fields the asymmetry persists but decreases. For the latter the distance to dose maximum is reduced by approximately 5 mm, the penumbra is increased by approximately 1 mm, and the 50% isodose line is shifted approximately 1 mm. The dose deposition in the presence of 1.5 T is affected, but the effect can be taken into account in a conventional treatment planning procedure. The impact of the altered dose deposition for clinical IMRT treatments is the topic of further research.

A six-bank multi-leaf system for high precision shaping of large fields.

In this study, we present the design for an alternative MLC system that allows high precision shaping of large fields. The MLC system consists of three layers of two opposing leaf banks. The layers are rotated 60 degrees relative to each other. The leaves in each bank have a standard width of 1 cm projected at the isocentre. Because of the symmetry of the collimator set-up it is expected that collimator rotation will not be required, thus simplifying the construction considerably. A 3D ray tracing computer program was developed in order to simulate the fluence profile for a given collimator and used to optimize the design and investigate its performance. The simulations show that a six-bank collimator will afford field shaping of fields of about 40 cm diameter with a precision comparable to that of existing mini MLCs with a leaf width of 4 mm.

Development of a regional hyperthermia treatment planning system.

A flexible and fast regional hyperthermia treatment planning system for the Coaxial TEM System has been devised and is presented. Using Hounsfield Unit based thresholding and manually outlining of the tumour, a 40 cm CT data set (slice thickness 5 mm) is segmented and down scaled to a resolution of 1 cm, requiring only 30 min. The SAR model is based on the finite-difference time-domain (FDTD) method. The number of time steps to achieve numerical stability has been determined and was found to be 7000. Various optimizations of the SAR model have been applied, resulting in a relatively short computation time of 3.7 h (memory requirements 121 MB) on a Pentium III, 450 MHz standard personal computer, running GNU/Linux. The model has been validated using absolute value(Ez) measurements in a standard phantom inserted in the Coaxial TEM Applicator under different conditions and a good agreement was found. Hyperthermia treatment planning in combination with the homemade visualization tools have provided much insight in the regional hyperthermia treatment with the Coaxial TEM Applicator.

Discretizing large traceable vessels and using DE-MRI perfusion maps yields numerical temperature contours that match the MR noninvasive measurements.

The success of hyperthermia treatments is dependent on thermal dose distribution. However, the three-dimensional temperature distribution remains largely unknown. Without this knowledge, the relationship between thermal dose and outcome is noisy, and therapy cannot be optimized. Accurate computations of thermal distribution can contribute to an optimized therapy. The hyperthermia modeling group in the Department of Radiotherapy, University Medical Center Utrecht devised a Discrete Vasculature [Kotte et al., Phys. Med. Biol. 41, 865-884 (1996)] model that accounts for the presence of vessel trees in the computational domain. The vessel tree geometry is tracked using magnetic resonance (MR) angiograms to a minimum diameter between 0.6 and 1 mm. However, smaller vessels (0.2-0.6 mm) are known to account for significant heat transfer. The hyperthermia group at Duke University Medical Center has proposed using perfusion maps derived from dynamic-enhanced magnetic resonance imaging to account for the tissue perfusion heterogeneity [Craciunescu et al., Int. J. Hyperthermia 17, 221-239 (2001)]. In addition, techniques for noninvasive temperature measurements have been devised to measure temperatures in vivo [Samulski et al., Int. J. Hypertherminal, 819-829 (1992)]. In this work, a patient with high-grade sarcoma has been retrospectively modeled to determine the temperature distribution achieved during a hyperthermia treatment. Available for this model were MR depicted geometry, angiograms, perfusion maps, as necessary for accurate thermal modeling, as well as MR thermometry data for validation purposes. The vasculature assembly through modifiable potential program [Van Leeuwen et al., IEEE Trans. Biomed. Eng. 45, 596-604 (1998)] was used in order to incorporate the traceable large vessels. Temperature simulations were made using different approaches to describe perfusion. The simulated cases were the bioheat equation with constant perfusion rates per tissue type, perfusion maps alone, tracked vessel tree and perfusion maps, and generated vessel tree. The results were compared with MR thermometry data for a single patient data set, concluding that a combination between large traceable vessels and perfusion map yields the best results for this particular patient. The technique has to be repeated on several patients, first with the same type of malignancy, and after that, on patients having malignancies at other different sites.

How to apply a discrete vessel model in thermal simulations when only incomplete vessel data are available.

For accurate predictions of the temperature distribution during hyperthermia treatment a thermal model should incorporate the individual impact of discrete vessels. In clinical practice not all vessels can be reconstructed individually. This paper investigates five possible strategies to model the thermal impact of these missing vessels. A tissue volume with a detailed, realistic, counter-current discrete vasculature is heated and the steady-state temperature distribution is calculated using our Discrete Vasculature (DIVA) thermal model. To mimic incomplete discrete vasculatures the full tree is gradually stripped, that is, the number of discretely described vessels is reduced in four steps until no discrete vessels are left. At each strip level the steady state temperature distribution is calculated for five different strategies to model the missing vessels. The strategies all use a local or global heat sink model in addition to the discrete vasculature. The resulting temperature distributions are compared with the full tree simulation. With increasing strip level the correspondence with the full tree simulation deteriorated for all strategies. An optimal strategy was found to model the missing vessels depending on the available angiographic data. It was also found that simulations with a decreased number of discrete vessels, or no vessels at all, yield temperatures which are too high. Theoretically this can be compensated by increasing the thermal conductivity; finding the optimal value is done empirically.

Temperature simulations in tissue with a realistic computer generated vessel network.

The practical use of a discrete vessel thermal model for hyperthermia treatment planning requires a number of choices with respect to the unknown part of the patient's vasculature. This work presents a study of the thermal effects of blood flow in a simple tissue geometry with a detailed artificial vessel network. The simulations presented here demonstrate that an incomplete discrete description of the detailed network results in a better prediction of the temperature distribution than is obtained using the conventional bio-heatsink equation. Therefore, efforts to obtain information on the positions of the large vessels in an individual hyperthermia patient will be rewarded with a more accurate prediction of the temperature distribution.

Calculation of change in brain temperatures due to exposure to a mobile phone.

In this study we evaluated for a realistic head model the 3D temperature rise induced by a mobile phone. This was done numerically with the consecutive use of an FDTD model to predict the absorbed electromagnetic power distribution, and a thermal model describing bioheat transfer both by conduction and by blood flow. We calculated a maximum rise in brain temperature of 0.11 degrees C for an antenna with an average emitted power of 0.25 W, the maximum value in common mobile phones, and indefinite exposure. Maximum temperature rise is at the skin. The power distributions were characterized by a maximum averaged SAR over an arbitrarily shaped 10 g volume of approximately 1.6 W kg(-1). Although these power distributions are not in compliance with all proposed safety standards, temperature rises are far too small to have lasting effects. We verified our simulations by measuring the skin temperature rise experimentally. Our simulation method can be instrumental in further development of safety standards.

Modelling the thermal impact of a discrete vessel tree.

Based on a modelling technique to calculate the thermal influence of a single vessel segment, a combination of segments representing a vessel tree is presented. At segment junctions the blood temperature is passed with a correction for a single vessel artefact. Blood leaving the modelled arterial vessel network at a junction or at the end of a terminal branch need not be equilibrated with the local surrounding tissue temperature. The thermal effect of this equilibration process can be taken into account using the 'sink set'. Blood entering a venous vessel tree is given an inflow temperature based on the tissue temperatures in the 'sample set'. With these sets we model the thermal impact of the vasculature too small to be taken into account discretely. The formation of the sink/sample sets is subject of current research; to show the capabilities of the presented method we present a minimal simulation collection with cube-shaped sets in combination with limited vasculature. The results of using the entire simulation volume as sink and sample sets for all the terminal branches matches the reference temperature profile best.

A flexible algorithm for construction of 3-D vessel networks for use in thermal modeling.

A new algorithm for the construction of artificial blood vessel networks is presented. The algorithm produces three-dimensional (3-D) geometrical representations of both arterial and venous networks. The key ingredient of the algorithm is a 3-D potential function defined in the tissue volume. This potential function controls the paths by which points are connected to existing vessels, thereby producing new vessel segments. The potential function has no physiological interpretation, but, by adjustment of parameters governing the potential, it is possible to produce networks that have physiologically meaningful geometrical properties. If desired, the veins can be generated counter current to the arteries. Furthermore, the potential function allows fashioning of the networks to the presence of bone or air cavities. The resulting networks can be used for thermal simulations of hyperthermia treatment.

Modelling tissue heating with ferromagnetic seeds.

Interstitial hyperthermia using ferromagnetic seeds demands accurate treatment planning: the seed characteristics and implant geometry must be determined prior to the treatment. A new, finite difference based, seed modelling method is presented. The seed, together with all its surrounding (non-tissue) layers is described as one unit, independent of the tissue grid. The calculation of the seed-tissue interaction is based on the local seed temperature and several tissue temperature samples in the direct vicinity. All the layers between the seed and the surrounding tissue are taken into account in this interaction calculation. The presented implementation describes the analytical solution of the modelled steady-state configurations very accurately. The separation between tissue and seed allows easy assessment of the resulting seed temperature profile which is essential to the optimization of the seed characteristics in treatment planning. The thermal effect due to blood flow in the modelled tissue volume surrounding the seed can be accounted for by inclusion of a heat sink term as well as by inclusion of realistic discrete vasculature.