• Home
• Preliminary Program
• News
• Speaker/Poster Information
• Conference Topics
• Programme Information
• Research Market / Beer & Wine party
• Awards
• Company Market
• Invited Speakers
• Important Dates
• Registration and Venue
• Authors Instructions
• Organizing Board
• Contact

13:00   Image Analysis - Cancer Chair: Chris de Korte 13:00 15 mins UWB SPIRAL ANTENNA DESIGN FOR BREAST TUMOR DETECTION Greet Baldewijns, Maarten Strackx, Patrick Reynaert, Paul Leroux Abstract: UWB imaging used for breast cancer detection aims to be a low-cost, low-hazard alternative for the X-ray mammography which is currently the most frequently used method for breast cancer screening. It works in a similar way to that of ground penetrating radar. All elements of an antenna array successively direct a short UWB pulse into the breast. The reflected signals are then detected by one or more receiving antennas. This sweep results in a set of received signals. A 3D image of the breast can be constructed based on this received set of signals. Malignant breast tissue is visible on the resulting 3D image [1]. The antenna design is a critical part of this detection method. An antenna was developed based on [2]. In [2] the antenna is optimized to radiate into air. The presented antenna however has been optimized to work efficiently when in contact with a matching medium with dielectric properties similar to breast tissue. The antenna consists of 2 spiral elements of the same size printed on an FR4 substrate with a thickness of 1.6 mm. On the other side of the substrate a ground plane with two circular slots is printed. Both spirals and slots are symmetrically placed with respect to the microstrip line. To achieve the desired bandwidth, the number of turns, width and distance between the spiral structures were modeled using CST Microwave Studio©. In order to facilitate the placement of the antenna against the breast a FR4 substrate was added on top of the ground plane. A metallic backplane was added to reflect the otherwise lost backward radiation into the matching medium. The backplane was placed at 8 mm from the feeding substrate. At this distance the interference of the reflected waves with the non-reflected waves has a positive effect on the radiation pattern. The width of the top substrate and the ideal location of the metallic backplane were both determined using parameter sweeps. An SMA connector was used to feed the antenna. In order to characterize the antenna, the S11 bandwidth and radiation patterns were simulated and measured. The relative radiation pattern was found by measuring the transmission between two identical spiral antennas in the desired measurement plane. The S11 parameter was measured with the top of the antenna immersed in a matching medium. The simulated and measured S11 parameter shows a -10 dB bandwidth from 3.8 to 15 GHz. Both the near and far field confirm that the antenna radiates into the matching medium. The measurements confirmed the simulated characteristics of the antenna. Both simulations and measurements prove the usability of the antenna for UWB breast cancer detection. 13:15 15 mins PATIENT-SPECIFIC FINITE ELEMENT MODELS DIFFERENTIATE BETWEEN PATIENTS WITH AND WITHOUT A PATHOLOGICAL FRACTURE IN METASTATIC BONE DISEASE Loes Derikx, Dennis Janssen, Yvette van der Linden, An Snyers, Nico Verdonschot, Esther Tanck Abstract: Introduction: Current clinical practice lacks an accurate predictor of the femoral fracture risk in patients suffering metastatic bone disease. This results in large numbers of patients who are over- and undertreated with complex surgery [1, 2]. The aim of this study is to assess whether patient-specific finite element (FE) models are able to discriminate between patients with an unexpected pathological fracture from patients who did not fracture their femur. Methods: Fourteen patients with painful femoral bone metastases with an expected low fracture risk were selected from a cohort of 66 patients. In nine cases patients suffered a uni- or bilateral fracture during follow-up (fracture group, F). Nine femora that did not fracture served as a control group (NF). Patient-specific FE models of these eighteen femora with proven malignancy were generated on the basis of quantitative computed tomography images of the femoral regions. We used four-noded tetrahedral elements and adopted non-linear isotropic material behaviour [3], based on phantom-calibrated grey values in the CT scans. The FE models were loaded until failure using an axial compressive load. The femoral failure load was calculated as the main outcome parameter. The failure location was defined by elements that plastically deformed when the failure load was reached. The mean failure load corrected for body weight, the work and the structural stiffness were compared between the fracture group and non-fracture group using independent T-tests. The predicted fracture locations were compared to clinical reports if available. Results: The FE-predicted failure load in the fracture group was significantly lower than in the non-fracture group (mean difference 3.66*BW, p < 0.01). No significant difference in structural stiffness was found. Good agreement was found between the predicted failure courses and the clinically reported fractures. Discussion: This study showed that patient-specific FE models improve upon current clinical methods in the prediction of fracture risk in metastatic bone disease. In a set of femora with a low fracture risk, FE models correctly differentiated between femora with and without an unexpected fracture by applying only a very simple loading condition. These results will be further validated in a larger patient population using more realistic loading conditions. 13:30 15 mins TUMOR LOCALIZATION AND TREATMENT ASSESSMENT IN DIFFUSE OPTICAL TOMOGRAPHY DURING NEO-ADJUVANT CHEMOTHERAPY IN BREAST CANCER Martijn van de Giessen, Bob Schaafsma, Judith Kroep, Martin Wasser, Cock van de Velde, Boudewijn Lelieveldt, Jouke Dijkstra, Alexander Vahrmeijer Abstract: Neo-adjuvant chemotherapy response may be monitored in a non-invasive way using multi-spectral diffuse optical tomography (DOT) [1]. A clinically feasible measurement method in these 3D volumes is proposed and compared to previously published whole breast measurements [2]. Twelve breast cancer patients were included, which received neo-adjuvant chemotherapy. Four DOT scans were obtained: before therapy, after 3 and 9 weeks, and pre-operative (18 weeks). Deoxyhemoglobin (Hb), oxyhemoglobin (HbO2) and percentages water and lipids were measured in volumes as measured in DCE-MRI scans before therapy, centered on areas of high scattering amplitude and approximate location in pre-treatment DCE-MRI and/or X-ray mammography. Patients were divided in responders and poor responders based on post-operative pathological assessment. Responders (R) and poor responders (PR) showed statistically significant (P<0.05) different development of Hb values at 3, 9 and 18 weeks within measurement volumes relative to the first DOT scan: 0.66/1.08 (R/PR, 3 weeks), 0.56/1.07 (9 weeks), 0.55/1.48 (18 weeks). For HbO2 results were similar, but only significant for 9 and 18 weeks: 0.71/1.03 (3 weeks), 0.53/0.89 (9 weeks), 0.56/1.06 (18 weeks). Percentages water and lipids were not significantly different. Whole breast measurements did not show statistically significant differences for Hb and HbO2 between responders and poor responders. Using the scatter amplitude to determine the tumor location in DOT volumes enables the measurement of Hb and HbO2 such that these parameters differ significantly between responders and poor responders. This potentially allows effectiveness assessment of neo-adjuvant treatment in breast cancer patients already at 3 weeks. REFERENCES [1] S. Srinivasan, B.W. Pogue, et al: Data subset algorithm for computationally efficient reconstruction of 3-D spectral imaging in diffuse optical tomography, Optics Express, 2006 [2] O. Falou, H. Soliman, et al.: Diffuse optical spectroscopy evaluation of treatment response in women with locally advanced breast cancer receiving neoadjuvant chemotherapy, Translational Oncology, 2012 13:45 15 mins CRITICAL TISSUES SEGMENTATION OF HEAD AND NECK CT IMAGES FOR HYPERTHERMIA TREATMENT PLANNING Valerio Fortunati, Rene Verhaart, Wiro Niessen, Jifke Veenland, Maarten Paulides, Gerard van Rhoon, Theo van Walsum Abstract: Introduction Optimization of hyperthermia treatment of head and neck tumors requires accurate treatment planning. Hyperthermia treatment planning (HTP) is based on tissue segmentation for 3D patient model generation. This process is currently performed manually, which is a tedious and time consuming procedure limiting the clinical applicability of this treatment. Methods We developed and evaluated a fully automatic multi atlas-based segmentation algorithm for the temperature sensitive (critical) tissues of head and neck CT images. These are: Cerebrum, Cerebellum, Brain-Stem, Myleum (Spinal-Cord) [1] and Eyeballs [2]. To overcome the large anatomical variability, multi atlas registration and intensity-based classification were combined. Each tissue is segmented independently using binary graph cut method [3]. A cost function composed of 1) an intensity energy term, 2) a spatial prior energy term based on the atlas registration and 3) a regularization term is globally minimized using graph cut. Afterwards the results from each tissue are combined. Experiments and results Data was collected from patients scheduled for combined radiotherapy and hyperthermia (HT) treatment. The data used in the evaluation study comprised 18 axial CT images scanned with the same protocol. These images were manually delineated from a medical radiation technologist student which was trained by radiation oncologist. Using this dataset the method was tuned and evaluated in a nested leave-one out experiment. The contribution of adding intensity based classification and regularization to atlas registration is evaluated comparing the graph-cut method (GC) with majority voting combination (MV) [4] of the multi atlas registration results. Both GC and MV results are evaluated with respect to manual delineations used as ground-truth. For the evaluation Dice similarity coefficient (DSC) and mean surface distances (MSD) were calculated. Combining intensity classification, regularization and multi atlas registration, the accuracy of the segmentation of Cerebrum, Cerebellum and Eyeball is significantly improved (tested using two sided, paired, Mann-Whitney test, p>0.05) with respect to MV. Overall a high correspondence was found for these tissues with a DSC median value higher than 0.87 and MSD lower than the voxel resolution in-plane (1mm). The method does not give a significant improving for Myleum and Brain-Stem segmentation accuracy. While the Myleum accuracy is comparable with the other tissues results in DSC and MSD, the Brain-Stem segmentation is less accurate. Both tissues are hard to be discerned from their surrounding relying on intensity information and this may be the cause of failure in accuracy improvement. Conclusion We are planning to investigate the inter-observer variability in method accuracy using different atlases for segmentation. Further the influences of the variation of accuracy on the SAR and temperature distribution prediction (and so on HTP) will be analyzed. The proposed method has potential to improve the efficiency of tissues delineation for HTP. REFERENCES [1] P. Sminia, J. van der Zees, J. Wondergem, J. Haveman, “Effect of hyperthermia on the central nervous system” Int. J. Hyperthermia, 10(1), 1-30. (1994). [2] J.A. Elder, “Ocular Effects of Radiofrequency Energy”, Bioelectromagnetics Supplement, 6, 148-161, (2003). [3] F. van der Lijn, T. den Heijer, M.M.B. Breteler, W.J. Niessen, “Hippocampus segmentation in MR images using atlas registration, voxel classification, and graph cuts” Neuroimage, 43, 708-720, (2008). [4] S. L. Hartmann, M. H. Parks, P. R. Martin, and B. M. Dawant, “Automatic segmentation of internal structures of the head in MR images using a combination of similarity and free-form transformations: Part II, validation on severely atrophied brains IEEE transactions on medical imaging 18, 917 (1999). 14:00 15 mins CFD CHALLENGE ON A PATIENT BASED GIANT ANEURSYM IN A CLINICAL SETTING Merih Cibis, Jolanda Wentzel, Frank Gijsen Abstract: Background: Aneurysms are local dilatations of the arterial walls and it is hypothesized that the development of an aneurysm is affected by hemodynamic parameters. CFD is generally used to determine intra-aneurysmal hemodynamics. Cebral [1] studied a patient based aneurysm with a proximal stenosis and applied CFD to predict the effect of stenosis removal. CFD is potentially useful to determine the flow data noninvasively, but is it really reliable? To answer this question, a challenge was organized to compare velocity and pressure in a patient based aneurysm model. Methods: We participated in the challenge in a clinical setting with limited resources and simulation time. We used 4 million tetrahedral elements which was the maximum allowable number of elements in our setting. Two unsteady simulations were performed (low flow and high flow) using realistic inflow waveforms. The steady simulations were at mean and peak flow rates for both unsteady inflows to check the influence of unsteadiness. Linear velocity and pressure basis functions were used which were enabled by using pressure stabilizing Petrov-Galerkin scheme. Segregated pressure based algorithm was chosen with backward Euler time integration. The convergence criterion was set to 1d-3. Results: Low flow rate steady simulations were fully converged. A narrow jet coming from the parent artery impinged to the aneurysm wall. The rest of the aneurysm was filled with low circulating flow. Main pressure drop was at the stenosis proximal to the aneurysm and at the neck of the aneurysm. The pressure drop at flow rate of 6.41 mL/sec was 7.8 mmHg. Steady and unsteady simulations at the same flow rates were in good agreement, so the influence of unsteadiness was found to be small. At high steady flow rates, fully resolved solution could not be achieved. Although the solution was smooth at the parent artery, numerical oscillations were observed at the jet region in the aneurysm sac. The pressure drop at peak flow rate of 11.42 mL/sec was 22.8 mmHg which is in good agreement with Cebral’s paper (25 mmHg). Our results were also consistent with the majority of the CFD groups participated in the challenge. Discussion: Performing simulations of intra-aneurysmal flow was difficult with limited resources due to high velocity gradients in the sac. Fully resolved solutions could not be reached at high flow rates. However, the results in the parent vessel were in good agreement with Cebral and other groups. Although not all details in flow structures in the sac were captured, the results showed that pressure drop computations are feasible with the settings of this study. Therefore, general conclusions were achieved with sufficiently accurate results in our clinical setting. Overall, CFD seems to be able to determine the pressure drop. REFERENCE [1] Cebral et al. Am J Neuroradiol 2011 Jan; 32:27-33 14:15 15 mins 3D ULTRASOUND-HISTOLOGY REGISTRATION FOR VALIDATION OF PROSTATE CANCER IMAGING METHODS Stefan Schalk, Tamerlan Saidov, Hessel Wijkstra, Massimo Mischi Abstract: Several ultrasound (US) methods for localization of PCa, like contrast-enhanced US imaging [1] and elastography [2], are emerging, but any of these methods need accurate validation prior to introduction into the clinic. The current gold standard for validation is histologic analysis of the prostate after excision, which requires accurate registration between histology slices and US images. This task is complicated by misalignment between histology and US data, as well as deformation of the prostate due to pressure from the transrectal US (TRUS) probe and to the fixation process after excision. Moreover, besides the prostate contours, no reliable landmarks are present in US. In this abstract, a solution to the validation issue is proposed by means of a dedicated 3D surface-based registration algorithm. To reconstruct the prostate shape, contours are manually drawn in a 2D TRUS transversal apex-to-base video. By correlating the mid area of the video with a sagittal image, the 3D location of the contours is estimated. In addition, a 3D reconstruction of the histology, including cancer location, is realized by aligning and interpolating 4-mm-thick prostate slices. From both prostate shapes, triangulated meshes are constructed. The two meshes are registered by a combination of an affine and an elastic transformation. Volumetric deformation is estimated by interpolation of the surface displacements without use of any internal landmark. Preliminary validation was performed by an in vitro experiment with four prostate-mimicking phantoms pierced by five elastic wires. The resulting average registration error was 1.6 mm, which is far below the histology slicing resolution of 4 mm. The method’s feasibility was tested in 12 human prostates for which we were able to overlay data from histopathologic analysis onto US imaging planes. In conclusion, the presented registration algorithm shows promising performance for accurate validation of US-based PCa imaging techniques. Future employment of the method for MRI validation or fusion can also be envisaged. REFERENCES [1] M.P.J. Kuenen, M. Mischi and H. Wijkstra, “Contrast-ultrasound diffusion imaging for localization of prostate cancer”, IEEE T-MI, Vol. 30(8), pp. 14931502, (2011). [2] M. Moradi, P. Mousavi and P. Abolmaesumi, “Computer-aided diagnosis of prostate cancer with emphasis on ultrasound-based approached: a review”, Ultrasound in Med. & Biol., Vol. 33(7), pp. 1010-1028, (2007).