U. Van der Heide

Label Fusion in Atlas-Based Segmentation Using a Selective and Iterative Method for Performance Level Estimation (SIMPLE)

In a multi-atlas based segmentation procedure, propagated atlas segmentations must be combined in a label fusion process. Some current methods deal with this problem by using atlas selection to construct an atlas set either prior to or after registration. Other methods estimate the performance of propagated segmentations and use this performance as a weight in the label fusion process. This paper proposes a selective and iterative method for performance level estimation (SIMPLE), which combines both strategies in an iterative procedure. In subsequent iterations the method refines both the estimated performance and the set of selected atlases. For a dataset of 100 MR images of prostate cancer patients, we show that the results of SIMPLE are significantly better than those of several existing methods, including the STAPLE method and variants of weighted majority voting.

Registration of Cervical MRI Using Multifeature Mutual Information

Radiation therapy for cervical cancer can benefit from image registration in several ways, for example by studying the motion of organs, or by (partially) automating the delineation of the target volume and other structures of interest. In this paper, the registration of cervical data is addressed using mutual information (MI) of not only image intensity, but also features that describe local image structure. Three aspects of the registration are addressed to make this approach feasible. First, instead of relying on a histogram-based estimation of mutual information, which poses problems for a larger number of features, a graph-based implementation of alpha-mutual information (alpha-MI) is employed. Second, the analytical derivative of alpha-MI is derived. This makes it possible to use a stochastic gradient descent method to solve the registration problem, which is substantially faster than non-derivative-based methods. Third, the feature space is reduced by means of a principal component analysis, which also decreases the registration time. The proposed technique is compared to a standard approach, based on the mutual information of image intensity only. Experiments are performed on 93 T2-weighted MR clinical data sets acquired from 19 patients with cervical cancer. Several characteristics of the proposed algorithm are studied on a subset of 19 image pairs (one pair per patient). On the remaining data (36 image pairs, one or two pairs per patient) the median overlap is shown to improve significantly compared to standard MI from 0.85 to 0.86 for the clinical target volume (CTV, p = 2 ldr10-2), from 0.75 to 0.81 for the bladder (p = 8 ldr 10-6), and from 0.76 to 0.77 for the rectum (p = 2 ldr 10-4). The registration error is improved at important tissue interfaces, such as that of the bladder with the CTV, and the interface of the rectum with the uterus and cervix.

SEGMENTATION OF THE PROSTATE IN MR IMAGES BY ATLAS MATCHING

Prostate cancer treatment by radiation therapy requires an accurate localisation of the prostate. For the treatment planning, primarily computed tomography (CT) images are used, but increasingly magnetic resonance (MR) images are added, because of their soft-tissue contrast. In current practice at our hospital, a manual delineation of the prostate is made, based on the CT and MR scans, which is a labour-intensive task. We propose an automatic segmentation method, based on non-rigid registration of a set of prelabelled MR atlas images. The algorithm consists of three stages. Firstly, the target image is nonrigidly registered with each atlas image, using mutual information as the similarity measure. After that, the best registered atlas images are selected by comparing the mutual information values after registration. Finally, the segmentation is obtained by averaging the selected deformed segmentations and thresholding the result. The method is evaluated on 22 images by calculating the overlap of automatic and manual segmentations. This results in a median Dice similarity coefficient of 0.82

Variation in target and rectum dose due to prostate deformation : an assessment by repeated MR imaging and treatment planning

In daily clinical practice, implanted fiducial markers are used to correct for prostate motion, but not for prostate deformation. The aim of this study is to investigate the variation in target and rectum dose due to the deformation of the prostate gland (without seminal vesicles). Therefore, we performed five to six MRI scans of eight healthy volunteers that exhibited large variation in rectal volume and thus prostate deformation. Prostate motion was corrected by a mask-based rigid registration which uses the delineation as well as the internal structures of the prostate gland. Per MRI scan, one IMRT plan with a PTV margin of 4 mm was created, resulting in 41 IMRT plans. The dose distribution of the IMRT plan based on the MRI scan with the minimum rectal volume was applied to the other rigidly registered MRI scans to evaluate the impact of prostate deformation. In conclusion, pre-treatment planning on the minimum rectal volume can cause a fraction dose increase (up to 15%) to the rectum due to prostate deformation. The impact on the total dose increase to the rectum depends on the intrapatient rectum variation during treatment, but is negligible with the currently used PTV margins in a fractionated treatment.

DPH is localised in two distinct populations in lipid vesicles

Abstract The fluorescence lifetime and anisotropy of DPH molecules in vesicles of lipids is commonly interpreted in terms of intrinsic structural heterogeneities of the bilayer. However, these heterogeneities are not revealed in experiments using other probe molecules. Here, we present an alternative analysis in terms of two distinct populations of DPH molecules in the bilayer structure. The results indicate that the heterogeneities seen by DPH reflect the variations of the packing density of the lipids in a direction perpendicular to the bilayer surface. The two populations are created by a competition between the interactions between the DPH molecules with the lipid chains and their tendency to access the regions of free volume in the middle of the bilayer.

Improving SNR and B1 transmit field for an endorectal coil in 7 T MRI and MRS of prostate cancer

Higher magnetic field strengths like 7 T and above are desirable for MR spectroscopy given the increased spectral resolution and signal to noise ratio. At these field strengths, substantial nonuniformities in B1+/− and radiofrequency power deposition become apparent. In this investigation, we propose an improvement on a conventionally used endorectal coil, through the addition of a second element (stripline). Both elements are used as transceivers. In the center of the prostate, approximately 40% signal to noise ratio increase is achieved. In fact, the signal to noise ratio gain obtained with the quadrature configuration locally can be even greater than 40% when compared to the single loop configuration. This is due to the natural asymmetry of the B1+/− fields at high frequencies, which causes destructive and constructive interference patterns. Global specific absorption rate is reduced by almost a factor of 2 as expected. Furthermore, approximately a 4‐fold decrease in local specific absorption rate is observed when normalized to the B1 values in the center of the prostate. Because of the 4‐fold local specific absorption rate decrease obtained with the dual channel setup for the same reference B1 value (20 μT at 3.5 cm depth into the prostate) as compared to the single loop, the transmission power B1 duty cycle can be increased by a factor 4. Consequently, when using the two‐element endorectal coil, the radiofrequency power deposition is significantly reduced and radiofrequency intense sequences with adiabatic pulses can be safely applied at 7 T for 1H magnetic resonance spectroscopy and MRI in the prostate. Altogether, in vivo 1H magnetic resonance spectroscopic imaging of prostate cancer with a fully adiabatic sequence operated at a minimum B1+ of 20 μT shows insensitivity to the nonuniform transmit field, while remaining within local specific absorption rate guidelines of 10 W/kg. Magn Reson Med, 2012.

Consequences of leaf calibration errors on IMRT delivery

IMRT treatments using multi-leaf collimators may involve a large number of segments in order to spare the organs at risk. When a large proportion of these segments are small, leaf positioning errors may become relevant and have therapeutic consequences. The performance of four head and neck IMRT treatments under eight different cases of leaf positioning errors has been studied. Systematic leaf pair offset errors in the range of +/-2.0 mm were introduced, thus modifying the segment sizes of the original IMRT plans. Thirty-six films were irradiated with the original and modified segments. The dose difference and the gamma index (with 2%/2 mm criteria) were used for evaluating the discrepancies between the irradiated films. The median dose differences were linearly related to the simulated leaf pair errors. In the worst case, a 2.0 mm error generated a median dose difference of 1.5%. Following the gamma analysis, two out of the 32 modified plans were not acceptable. In conclusion, small systematic leaf bank positioning errors have a measurable impact on the delivered dose and may have consequences for the therapeutic outcome of IMRT.

Influence of the linac design on intensity-modulated radiotherapy of head-and-neck plans

In this study, we quantify the impact of linac/MLC design parameters on IMRT treatment plans. The investigated parameters were leaf width in the MLC, leaf transmission, related to the thickness of the leaves, and penumbra related primarily to the source size. Seven head-and-neck patients with stage T1-T3N0-N2cM0 oropharyngeal cancer were studied. For each patient nine plans were made with a different set of linac/MLC parameters. The plans were optimized in Pinnacle(3) v7.6c and PLATO RTS v2.6.4, ITP v1.1.8. A hypothetical ideal linac/MLC was introduced to investigate the influence of one parameter at a time without interaction of other parameters. When any of the three parameters was increased from the ideal set-up values (leaf width 2.5 mm, transmission 0%, penumbra 3 mm), the mean dose to the parotid glands increased, given the same tumour coverage. The largest increase was found for increasing leaf transmission. The investigation showed that by changing more than one parameter of the ideal linac/MLC set-up, the increase in the mean dose was smaller than the sum of dose increments for each parameter separately. As a reference to clinical practice, we also optimized the plans of the seven patients with the clinically used Elekta SLi 15, equipped with a standard MLC with a leaf width of 10 mm. As compared to the ideal linac, this resulted in an increase of the average dose to the parotid glands of 5.8 Gy.

SPECTROSCOPIC AND ORIENTATIONAL PROPERTIES OF CHLOROPHYLL a AND CHLOROPHYLL b IN LIPID MEMBRANES

Abstract

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.