Table of contents

It is well-known that in pinhole SPECT (single-photon-emission computed tomography), iterative reconstruction methods including accurate estimations of the system response matrix can lead to submillimeter spatial resolution. There are two different methods for obtaining the system response matrix: those that model the system analytically using an approach including an experimental characterization of the detector response, and those that make use of Monte Carlo simulations. Methods based on analytical approaches are faster and handle the statistical noise better than those based on Monte Carlo simulations, but they require tedious experimental measurements of the detector response. One suggested approach for avoiding an experimental characterization, circumventing the problem of statistical noise introduced by Monte Carlo simulations, is to perform an analytical computation of the system response matrix combined with a Monte Carlo characterization of the detector response. Our findings showed that this approach can achieve high spatial resolution similar to that obtained when the system response matrix computation includes an experimental characterization. Furthermore, we have shown that using simulated detector responses has the advantage of yielding a precise estimate of the shift between the point of entry of the photon beam into the detector and the point of interaction inside the detector. Considering this, it was possible to slightly improve the spatial resolution in the edge of the field of view.

The efficacy of Positron Emission Tomography (PET) imaging relies fundamentally on the ability of the system to accurately identify true coincidence events. With existing systems, this is currently accomplished with an energy acceptance criterion followed by correction techniques to remove suspected false coincidence events. These corrections generally result in signal and contrast loss and thus limit the PET system's ability to achieve optimum image quality. A key property of annihilation radiation is that the photons are polarised with respect to each other. This polarisation correlation offers a potentially powerful discriminator, independent of energy, to accurately identify true events. In this proof of concept study, we investigate how photon polarisation information can be exploited in PET imaging by developing a method to discriminate true coincidences using the polarisation correlation of annihilation pairs. We implement this method using a Geant4 PET simulation of a GE Advance/Discovery LS system and demonstrate the potential advantages of the polarisation coincidence selection method over a standard energy criterion method. Current PET ring detectors are not capable of exploiting the polarisation correlation of the photon pairs. Compton PET systems, however are promising candidates for this application. We demonstrate the feasibility of a two-component Compton camera system in identifying true coincidences with Monte Carlo simulations. Our study demonstrates the potential of improving signal gain using polarisation, particularly for high photon emission rates. We also demonstrate the ability of the Compton camera at exploiting this polarisation correlation in PET.

This paper describes the modelisation of the Elekta XVI Cone Beam Computed Tomography (CBCT) machine components with Geant4 and its validation against calibration data taken for two commonly used machine setups. Preliminary dose maps of simulated CBCTs coming from this modelisation work are presented. This study is the first step of a research project, GHOST, aiming to improve the understanding of late toxicity risk in external beam radiotherapy patients by simulating dose depositions integrated from different sources (imaging, treatment beam) over the entire treatment plan. The second cancer risk will then be derived from different models relating irradiation dose and second cancer risk.

This work presents a Monte Carlo model of Leksell Gamma Knife Perfexion as well as the main parameters of the dose distribution in the standard phantom obtained using this model. The model is developed in the Geant4 simulation toolkit in a modular way which enables its reuse in other Perfexion studies. Large phase space files were created, containing particles that are entering the inner machine cavity after being transported through the collimation system. All 14 output factors of the machine and effective output factors for both the 4 mm (0.830 ± 0.009) and 8 mm (0.921 ± 0.004) collimators were calculated. Dose profiles along the main axes are also included for each collimator size. All results are compared to the values obtained from the treatment planning system, from experiments, and from other Monte Carlo models.

The accuracy of radiopharmaceutical absorbed dose distributions computed through Monte Carlo (MC) simulations is mostly limited by the low spatial resolution of 3D imaging techniques used to define the simulation geometry. This issue also persists with the implementation of realistic hybrid models built using polygonal mesh and/or NURBS as they require to be simulated in their voxel form in order to reduce computation times. The existing trade-off between voxel size and simulation speed leads on one side, in an overestimation of the size of small radiosensitive structures such as the skin or hollow organs walls and, on the other, to unnecessarily detailed voxelization of large, homogeneous structures.

We developed a set of computational tools based on VTK and Geant4 in order to build multi-resolution organ models. Our aim is to use different voxel sizes to represent anatomical regions of different clinical relevance: the MC implementation of these models is expected to improve spatial resolution in specific anatomical structures without significantly affecting simulation speed. Here we present the tools developed through a proof of principle example. Our approach is validated against the standard Geant4 technique for the simulation of voxel geometries.

When a carbon beam interacts with human tissues, many secondary fragments are produced into the tumor region and the surrounding healthy tissues. Therefore, in hadrontherapy precise dose calculations require Monte Carlo tools equipped with complex nuclear reaction models. To get realistic predictions, however, simulation codes must be validated against experimental results; the wider the dataset is, the more the models are finely tuned.

Since no fragmentation data for tissue-equivalent materials at Fermi energies are available in literature, we measured secondary fragments produced by the interaction of a 55.6 MeV u⁻¹¹²C beam with thick muscle and cortical bone targets. Three reaction models used by the Geant4 Monte Carlo code, the Binary Light Ions Cascade, the Quantum Molecular Dynamic and the Liege Intranuclear Cascade, have been benchmarked against the collected data. In this work we present the experimental results and we discuss the predictive power of the above mentioned models.

Hadrontherapy is an innovative radiation therapy modality for which one of the main key advantages is the target conformality allowed by the physical properties of ion species. However, in order to maximise the exploitation of its potentialities, online monitoring is required in order to assert the treatment quality, namely monitoring devices relying on the detection of secondary radiations. Herein is presented a method based on Monte Carlo simulations to optimise a multi-slit collimated camera employing time-of-flight selection of prompt-gamma rays to be used in a clinical scenario. In addition, an analytical tool is developed based on the Monte Carlo data to predict the expected precision for a given geometrical configuration. Such a method follows the clinical workflow requirements to simultaneously have a solution that is relatively accurate and fast. Two different camera designs are proposed, considering different endpoints based on the trade-off between camera detection efficiency and spatial resolution to be used in a proton therapy treatment with active dose delivery and assuming a homogeneous target.

Gold nanoparticles (GNPs) have shown potential to be used as a radiosensitizer for radiation therapy. Despite extensive research activity to study GNP radiosensitization using photon beams, only a few studies have been carried out using proton beams. In this work Monte Carlo simulations were used to assess the dose enhancement of GNPs for proton therapy. The enhancement effect was compared between a clinical proton spectrum, a clinical 6 MV photon spectrum, and a kilovoltage photon source similar to those used in many radiobiology lab settings. We showed that the mechanism by which GNPs can lead to dose enhancements in radiation therapy differs when comparing photon and proton radiation. The GNP dose enhancement using protons can be up to 14 and is independent of proton energy, while the dose enhancement is highly dependent on the photon energy used. For the same amount of energy absorbed in the GNP, interactions with protons, kVp photons and MV photons produce similar doses within several nanometers of the GNP surface, and differences are below 15% for the first 10 nm. However, secondary electrons produced by kilovoltage photons have the longest range in water as compared to protons and MV photons, e.g. they cause a dose enhancement 20 times higher than the one caused by protons 10 μm away from the GNP surface. We conclude that GNPs have the potential to enhance radiation therapy depending on the type of radiation source. Proton therapy can be enhanced significantly only if the GNPs are in close proximity to the biological target.

In this study, fragmentation yields of carbon therapy beams are estimated using the Geant4 simulation toolkit version 9.5. Simulations are carried out in a step-by-step mode using the Geant4-DNA processes for each of the major contributing fragments. The energy of the initial beam is taken 400 MeV amu⁻¹ as this is the highest energy, which is used for medical accelerators and this would show the integral role of secondary contributions in radiotherapy irradiations. The obtained results showed that 64% of the global dose deposition is initiated by carbon ions, while up to 36% is initiated by the produced fragments including all their isotopes. The energy deposition clustering yields of each of the simulated fragments are then estimated using the DBSCAN clustering algorithm and they are compared to the yields of the incident primary beam.

We propose the split exponential track length estimator (seTLE), a new kerma-based method combining the exponential variant of the TLE and a splitting strategy to speed up Monte Carlo (MC) dose computation for low energy photon beams. The splitting strategy is applied to both the primary and the secondary emitted photons, triggered by either the MC events generator for primaries or the photon interactions generator for secondaries. Split photons are replaced by virtual particles for fast dose calculation using the exponential TLE. Virtual particles are propagated by ray-tracing in voxelized volumes and by conventional MC navigation elsewhere. Hence, the contribution of volumes such as collimators, treatment couch and holding devices can be taken into account in the dose calculation.

We evaluated and analysed the seTLE method for two realistic small animal radiotherapy treatment plans. The effect of the kerma approximation, i.e. the complete deactivation of electron transport, was investigated. The efficiency of seTLE against splitting multiplicities was also studied. A benchmark with analog MC and TLE was carried out in terms of dose convergence and efficiency.

The results showed that the deactivation of electrons impacts the dose at the water/bone interface in high dose regions. The maximum and mean dose differences normalized to the dose at the isocenter were, respectively of 14% and 2% . Optimal splitting multiplicities were found to be around 300. In all situations, discrepancies in integral dose were below 0.5% and 99.8% of the voxels fulfilled a 1%/0.3 mm gamma index criterion. Efficiency gains of seTLE varied from 3.2 × 10⁵ to 7.7 × 10⁵ compared to analog MC and from 13 to 15 compared to conventional TLE.

In conclusion, seTLE provides results similar to the TLE while increasing the efficiency by a factor between 13 and 15, which makes it particularly well-suited to typical small animal radiation therapy applications.

Changes in the viscoelastic parameters of soft biological tissues often correlate with progression of disease, trauma or injury, and response to treatment. Identifying the most appropriate viscoelastic model, then estimating and monitoring the corresponding parameters of that model can improve insight into the underlying tissue structural changes. MR Elastography (MRE) provides a quantitative method of measuring tissue viscoelasticity. In a previous study by the authors (Yasar et al2013Magn. Reson. Med. 70 479–89), a silicone-based phantom material was examined over the frequency range of 200 Hz–7.75 kHz using MRE, an unprecedented bandwidth at that time. Six viscoelastic models including four integer order models and two fractional order models, were fit to the wideband viscoelastic data (measured storage and loss moduli as a function of frequency). The 'fractional Voigt' model (spring and springpot in parallel) exhibited the best fit and was even able to fit the entire frequency band well when it was identified based only on a small portion of the band. This paper is an extension of that study with a wider frequency range from 500 Hz to 16 kHz. Furthermore, more fractional order viscoelastic models are added to the comparison pool. It is found that added complexity of the viscoelastic model provides only marginal improvement over the 'fractional Voigt' model. And, again, the fractional order models show significant improvement over integer order viscoelastic models that have as many or more fitting parameters.

Ultrasound radiation force-based methods can quantitatively evaluate tissue viscoelastic material properties. One of the limitations of the current methods is neglecting the inherent anisotropy nature of certain tissues. To explore the phenomenon of anisotropy in a laboratory setting, we created two phantom designs incorporating fibrous and fishing line material with preferential orientations. Four phantoms were made in a cube-shaped mold; both designs were arranged in multiple layers and embedded in porcine gelatin using two different concentrations (8%, 14%). An excised sample of pork tenderloin was also studied. Measurements were made in the phantoms and the pork muscle at different angles by rotating the phantom with respect to the transducer, where 0° and 180° were defined along the fibers, and 90° and 270° across the fibers. Shear waves were generated and measured by a Verasonics ultrasound system equipped with a linear array transducer. For the fibrous phantom, the mean and standard deviations of the shear wave speeds along (0°) and across the fibers (90°) with 8% gelatin were 3.60  ±  0.03 and 3.18  ±  0.12 m s⁻¹ and with 14% gelatin were 4.10  ±  0.11 and 3.90  ±  0.02 m s⁻¹. For the fishing line material phantom, the mean and standard deviations of the shear wave speeds along (0°) and across the fibers (90°) with 8% gelatin were 2.86  ±  0.20 and 2.44  ±  0.24 m s⁻¹ and with 14% gelatin were 3.40  ±  0.09 and 2.84  ±  0.14 m s⁻¹. For the pork muscle, the mean and standard deviations of the shear wave speeds along the fibers (0°) at two different locations were 3.83  ±  0.16 and 3.86  ±  0.12 m s⁻¹ and across the fibers (90°) were 2.73  ±  0.18 and 2.70  ±  0.16 m s⁻¹, respectively. The fibrous and fishing line gelatin-based phantoms exhibited anisotropy that resembles that observed in the pork muscle.

This study investigated the accuracy of positioning and irradiation targeting for multiple off-isocenter targets in intracranial image-guided radiation therapy (IGRT). A phantom with nine circular targets was created to evaluate both accuracies. First, the central point of the isocenter target was positioned with a combination of an ExacTrac x-ray (ETX) and a 6D couch. The positioning accuracy was determined from the deviations of coordinates of the central point in each target obtained from the kV-cone beam computed tomography (kV-CBCT) for IGRT and the planning CT. Similarly, the irradiation targeting accuracy was evaluated from the deviations of the coordinates between the central point of each target and the central point of each multi-leaf collimator (MLC) field for multiple targets. Secondly, the 6D couch was intentionally rotated together with both roll and pitch angles of 0.5° and 1° at the isocenter and similarly the deviations were evaluated. The positioning accuracy for all targets was less than 1 mm after 6D positioning corrections. The irradiation targeting accuracy was up to 1.3 mm in the anteroposterior (AP) direction for a target 87 mm away from isocenter. For the 6D couch rotations with both roll and pitch angles of 0.5° and 1°, the positioning accuracy was up to 1.0 mm and 2.3 mm in the AP direction for the target 87 mm away from the isocenter, respectively. The irradiation targeting accuracy was up to 2.1 mm and 2.6 mm in the AP direction for the target 87 mm away from the isocenter, respectively. The off-isocenter irradiation targeting accuracy became worse than the positioning accuracy. Both off-isocenter accuracies worsened in proportion to rotation angles and the distance from the isocenter to the targets. It is necessary to examine the set-up margin for off-isocenter multiple targets at each institution because irradiation targeting accuracy is peculiar to the linac machine.

In this article, a set of three computer vision tools, including scale invariant feature transform (SIFT), a measure of focus, and a measure based on tractography are demonstrated to be useful in replacing the eye of the expert in the optimization of the reconstruction parameters in x-ray in-line phase tomography. We demonstrate how these computer vision tools can be used to inject priors on the shape and scale of the object to be reconstructed. This is illustrated with the Paganin single intensity image phase retrieval algorithm in heterogeneous soft tissues of biomedical interest, where the selection of the reconstruction parameters was previously made from visual inspection or physical assumptions on the composition of the sample.

To solve the multicollinearity issue and unequal contribution of vascular parameters for the quantification of angiogenesis, we developed a quantification evaluation method of vascular parameters for angiogenesis based on in vivo micro-CT imaging of hindlimb ischemic model mice. Taking vascular volume as the ground truth parameter, nine vascular parameters were first assembled into sparse principal components (PCs) to reduce the multicolinearity issue. Aggregated boosted trees (ABTs) were then employed to analyze the importance of vascular parameters for the quantification of angiogenesis via the loadings of sparse PCs. The results demonstrated that vascular volume was mainly characterized by vascular area, vascular junction, connectivity density, segment number and vascular length, which indicated they were the key vascular parameters for the quantification of angiogenesis. The proposed quantitative evaluation method was compared with both the ABTs directly using the nine vascular parameters and Pearson correlation, which were consistent. In contrast to the ABTs directly using the vascular parameters, the proposed method can select all the key vascular parameters simultaneously, because all the key vascular parameters were assembled into the sparse PCs with the highest relative importance.

Tumour tracking with scanned particle beams potentially requires accurate 3D information on both tumour motion and related density variations. We have previously developed a model-based motion reconstruction method, which allows for the prediction of deformable motions from sparsely sampled surrogate motions tracked via an on-board imaging system (Zhang et al (2013 Phys. Med. Biol.58 8621)). Here, we investigate the potential effectiveness of tumour tracking for scanned proton beam therapy using such an approach to guide scanned beam tracking, together with the effectiveness of 're-tracking' for reducing residual motion effects due to tracking uncertainties. Three different beam tracking strategies (2D, 2D deformable and 3D) have been applied to three different liver motion cases, with mean magnitudes ranging from 10–20 mm. All simulations have been performed using simulated 4DCTs derived from 4DMRI datasets, whereby inter-breath-cycle motion variability is taken into account. The results show that, without beam tracking, large interplay effects are observed for all motion cases, resulting in CTV D_5–95 values of 34.9/58.5/79.4% for the three cases, respectively. These can be reduced to 16.9/18.8/29.1% with 2D tracking, to 15.5/17.9/23.3% with 2D deformable tracking and to 15.1/17.8/21.0% with 3D tracking. Clear 'inverse interplay' effects have also been observed in the proximal portion of the field. However, with three-times re-tracking, D_5–95 for the largest motions (20 mm) can be reduced to 13.0/12.8% for 2D and 3D tracking, respectively, and 'hot spots' resulting from the 'inverse interplay' effect can be substantially reduced. In summary, we have found that, for motions over 10 mm, tracking alone cannot fully mitigate motion effects, and can lead to substantially increased doses to normal tissues in the entrance path of the field. However, three-times re-tracking substantially improves the effectiveness of all types of beam tracking, with substantial advantages of 3D over 2D re-tracking only being observed for the largest motion scenario investigated.

Quantitative computed tomography (QCT) is increasingly used in osteoporosis studies to assess volumetric bone mineral density (vBMD), bone quality and strength. However, QCT is confronted by technical issues in the clinical research setting, such as potentially confounding effects of body size on vBMD measurements and lack of standard approaches to scanner cross-calibration, which affects measurements of vBMD in multicenter settings. In this study, we addressed systematic inter-scanner differences and subject-dependent body size errors using a novel anthropomorphic hip phantom, containing a calibration hip to estimate correction equations, and a contralateral test hip to assess the quality of the correction. We scanned this phantom on four different scanners and we applied phantom-derived corrections to in vivo images of 16 postmenopausal women scanned on two scanners. From the phantom study, we found that vBMD decreased with increasing phantom size in three of four scanners and that inter-scanner variations increased with increasing phantom size. In the in vivo study, we observed that inter-scanner corrections reduced systematic inter-scanner mean vBMD differences but that the inter-scanner precision error was still larger than expected from known intra-scanner precision measurements. In conclusion, inter-scanner corrections and body size influence should be considered when measuring vBMD from QCT images.

This paper describes a block detector multiplexing technique that simplifies the extension of axial length or ring radius without employing extra data acquisition channels. The proposed multiplexing circuit (bipolar multiplexing) multiplexes block detectors by encoding the position-related signal outputs with different combinations of polarity. Accordingly, it is possible to distinguish the detectors using a shared readout channel. This method was evaluated by assessing one to 16 block detectors using four data acquisition channels and one trigger input at different count rates. The experimental results showed that the multiplexing of block detectors did not significantly degrade timing, energy or spatial performance at low count rates, while reducing the required number of data acquisition channels and wire routing density. On the other hand, the degradation was seen as total count rates of all blocks increased.

Computational medicine aims at employing personalised computational models in diagnosis and treatment planning. The use of such models to help physicians in finding the best treatment for low back pain (LBP) is becoming popular. One of the challenges of creating such models is to derive patient-specific anatomical and tissue models of the lumbar intervertebral discs (IVDs), as a prior step. This article presents a segmentation scheme that obtains accurate results irrespective of the degree of IVD degeneration, including pathological discs with protrusion or herniation. The segmentation algorithm, employing a novel feature selector, iteratively deforms an initial shape, which is projected into a statistical shape model space at first and then, into a B-Spline space to improve accuracy.

The method was tested on a MR dataset of 59 patients suffering from LBP. The images follow a standard T2-weighted protocol in coronal and sagittal acquisitions. These two image volumes were fused in order to overcome large inter-slice spacing. The agreement between expert-delineated structures, used here as gold-standard, and our automatic segmentation was evaluated using Dice Similarity Index and surface-to-surface distances, obtaining a mean error of 0.68 mm in the annulus segmentation and 1.88 mm in the nucleus, which are the best results with respect to the image resolution in the current literature.

Many applications in medical imaging include image registration for matching of images from the same or different modalities. In the case of full data sampling, the respective reconstructed images are usually of such a good image quality that standard deformable volume-to-volume (3D–3D) registration approaches can be applied. But research in temporal-correlated image reconstruction and dose reductions increases the number of cases where rawdata are available from only few projection angles. Here, deteriorated image quality leads to non-acceptable deformable volume-to-volume registration results. Therefore a registration approach is required that is robust against a decreasing number of projections defining the target position. We propose a deformable volume-to-rawdata (3D–2D) registration method that aims at finding a displacement vector field maximizing the alignment of a CT volume and the acquired rawdata based on the sum of squared differences in rawdata domain. The registration is constrained by a regularization term in accordance with a fluid-based diffusion. Both cost function components, the rawdata fidelity and the regularization term, are optimized in an alternating manner. The matching criterion is optimized by a conjugate gradient descent for nonlinear functions, while the regularization is realized by convolution of the vector fields with Gaussian kernels. We validate the proposed method and compare it to the demons algorithm, a well-known 3D–3D registration method. The comparison is done for a range of 4–60 target projections using datasets from low dose tomographic fluoroscopy as an application example. The results show a high correlation to the ground truth target position without introducing artifacts even in the case of very few projections. In particular the matching in the rawdata domain is improved compared to the 3D–3D registration for the investigated range. The proposed volume-to-rawdata registration increases the robustness regarding sparse rawdata and provides more stable results than volume-to-volume approaches. By applying the proposed registration approach to low dose tomographic fluoroscopy it is possible to improve the temporal resolution and thus to increase the robustness of low dose tomographic fluoroscopy.

A heterogeneous distribution of radionuclides emitting low-energy electrons in the testicles may result in a significant difference between an absorbed dose to the radiosensitive spermatogonia and the mean absorbed dose to the whole testis. This study focused on absorbed dose distribution in patients at a finer scale than normally available in clinical dosimetry, which was accomplished by combining a small-scale dosimetry model with patient pharmacokinetic data. The activity in the testes was measured and blood sampling was performed for patients that underwent pre-therapy imaging with ¹¹¹In-Zevalin^®. Using compartment modeling, testicular activity was separated into two components: vascular and extravascular. The uncertainty of absorbed dose due to geometry variations between testicles was explored by an assumed activity micro-distribution and by varying the radius of the interstitial tubule. Results showed that the absorbed dose to germ cells might be strongly dependent on the location of the radioactive source, and may exceed the absorbed dose to the whole testis by as much as a factor of two. Small-scale dosimetry combined with compartmental analysis of clinical data proved useful for gauging tissue dosimetry and interpreting how intrinsic geometric variation influences the absorbed dose.

A novel approach to proton CT reconstruction using backprojection-then-filtering (BPF) is proposed. A list-mode algorithm is formulated accommodating non-linear proton paths. The analytical form is derived for the deblurring kernel necessary for the filtering step. Further, a finite matrix correction is derived to correct for the limited size of the backprojection matrix. High quantitative accuracy in relative stopping power is demonstrated (⩽0.1%) using Monte Carlo simulations. This accuracy makes the algorithm a promising candidate for future proton CT systems in proton therapy applications. For the purposes of reconstruction, each proton path in the object-of-interest was estimated based on a cubic spline fit to the proton entry and exit vectors. The superior spatial-resolution of the BPF method over the standard filtering-then-backprojection approach is demonstrated. As the BPF algorithm requires only one backprojection and filtering operation on a scan data set, it also offers computational advantages over an iterative reconstruction approach.

In GPU-based Monte Carlo simulations for radiotherapy dose calculation, source modelling from a phase-space source can be an efficiency bottleneck. Previously, this has been addressed using phase-space-let (PSL) sources, which provided significant efficiency enhancement. We propose that additional speed-up can be achieved through the use of a hybrid primary photon point source model combined with a secondary PSL source. A novel phase-space derived and histogram-based implementation of this model has been integrated into gDPM v3.0. Additionally, a simple method for approximately deriving target photon source characteristics from a phase-space that does not contain inheritable particle history variables (LATCH) has been demonstrated to succeed in selecting over 99% of the true target photons with only ~0.3% contamination (for a Varian 21EX 18 MV machine). The hybrid source model was tested using an array of open fields for various Varian 21EX and TrueBeam energies, and all cases achieved greater than 97% chi-test agreement (the mean was 99%) above the 2% isodose with 1% / 1 mm criteria. The root mean square deviations (RMSDs) were less than 1%, with a mean of 0.5%, and the source generation time was 4–5 times faster. A seven-field intensity modulated radiation therapy patient treatment achieved 95% chi-test agreement above the 10% isodose with 1% / 1 mm criteria, 99.8% for 2% / 2 mm, a RMSD of 0.8%, and source generation speed-up factor of 2.5.

3ʹ-(¹⁸F)fluoro-3ʹ-deoxy-L-thymidine (FLT) is a PET marker of cellular proliferation. Its tissue uptake rate is often quantified with a Standardized Uptake Value (SUV), although kinetic analysis provides a more accurate quantification. The purpose of this study is to investigate the heterogeneity in FLT stabilization phenomena.

The study was done on 15 canines with spontaneously occurring sinonasal tumours. They were imaged dynamically for 90 min with FLT PET/CT twice; before and during the radiotherapy. Images were analyzed for kinetics on a voxel basis through compartmental analysis. Stabilization curves were calculated as a time-dependant correlation between the time-dependant SUV and the kinetic parameters (voxel values within the tumour were correlated). Stabilization curves were analyzed for stabilization speed, maximal correlation and correlation decrease following the maximal correlation. These stabilization parameters were correlated with the region-averaged kinetic parameters.

The FLT SUV was highly correlated with vasculature fraction immediately post-injection, followed by maximum in correlation with the perfusion/permeability. At later times post-injection the FLT SUV was highly correlated (Pearson correlation coefficient above 0.95) with the FLT influx parameter for cases with tumour-averaged SUV_{30–50 min} above 2, while others were indeterminate (correlation coefficients from 0.1 to 0.97). All cases with highly correlated SUV and FLT influx parameter had correlation coefficient within 0.5% of its maximum in the period of 30–50 min post-injection. Stabilization time was inversely proportional to the FLT influx rate. Correlation between the FLT SUV and FLT influx parameter dropped at later times post-injection with drop being proportional to the dephosphorylation rate. The FLT was found to be metabolically stable in canines.

FLT PET imaging protocol should define minimal and maximal FLT uptake period, which would be 30–50 min for our patients. Additionally, kinetic analysis should be used when low FLT avidity is expected. Low SUVs should be treated with great caution.

The conversion coefficients (CCs) relate protection quantities, mean absorbed dose (D_T) and effective dose (E), with physical radiation field quantities, such as fluence (Φ). The calculation of CCs through Monte Carlo simulations is useful for estimating the dose in individuals exposed to radiation. The aim of this work was the calculation of conversion coefficients for absorbed and effective doses per fluence (D_T/ Φ and E/Φ) using a sitting and standing female hybrid phantom (UFH/NCI) exposure to monoenergetic protons with energy ranging from 2 MeV to 10 GeV. The radiation transport code MCNPX was used to develop exposure scenarios implementing the female UFH/NCI phantom in sitting and standing postures. Whole-body irradiations were performed using the recommended irradiation geometries by ICRP publication 116 (AP, PA, RLAT, LLAT, ROT and ISO). In most organs, the conversion coefficients D_T/Φ were similar for both postures. However, relative differences were significant for organs located in the abdominal region, such as ovaries, uterus and urinary bladder, especially in the AP, RLAT and LLAT geometries. Anatomical differences caused by changing the posture of the female UFH/NCI phantom led an attenuation of incident protons with energies below 150 MeV by the thigh of the phantom in the sitting posture, for the front-to-back irradiation, and by the arms and hands of the phantom in the standing posture, for the lateral irradiation.