S. Zwick

Publications of S. Zwick

• Fully automated intrinsic respiratory and cardiac gating for small animal CT.

  Authors: J Kuntz, J Dinkel, S Zwick, T Bäuerle, M Grasruck, F Kiessling, R Gupta, W Semmler, S H Bartling

  Physics in medicine and biology. 04/2010; 55(7):2069-85.

  A fully automated, intrinsic gating algorithm for small animal cone-beam CT is described and evaluated. A parameter representing the organ motion, derived from the raw projection images, is used forA fully automated, intrinsic gating algorithm for small animal cone-beam CT is described and evaluated. A parameter representing the organ motion, derived from the raw projection images, is used for both cardiac and respiratory gating. The proposed algorithm makes it possible to reconstruct motion-corrected still images as well as to generate four-dimensional (4D) datasets representing the cardiac and pulmonary anatomy of free-breathing animals without the use of electrocardiogram (ECG) or respiratory sensors. Variation analysis of projections from several rotations is used to place a region of interest (ROI) on the diaphragm. The ROI is cranially extended to include the heart. The centre of mass (COM) variation within this ROI, the filtered frequency response and the local maxima are used to derive a binary motion-gating parameter for phase-sensitive gated reconstruction. This algorithm was implemented on a flat-panel-based cone-beam CT scanner and evaluated using a moving phantom and animal scans (seven rats and eight mice). Volumes were determined using a semiautomatic segmentation. In all cases robust gating signals could be obtained. The maximum volume error in phantom studies was less than 6%. By utilizing extrinsic gating via externally placed cardiac and respiratory sensors, the functional parameters (e.g. cardiac ejection fraction) and image quality were equivalent to this current gold standard. This algorithm obviates the necessity of both gating hardware and user interaction. The simplicity of the proposed algorithm enables adoption in a wide range of small animal cone-beam CT scanners.

• Simulation-based comparison of two approaches frequently used for dynamic contrast-enhanced MRI

  Authors: S. Zwick, G Brix, P S Tofts, R. Strecker, A. Kopp-Schneider, H. Laue, W Semmler, F Kiessling

  Fraunhofer MEVIS.

  The purpose was to compare two approaches for the acquisition and analysis of dynamic-contrast-enhanced MRI data with respect to differences in the modelling of the arterial input-function (AIF), theThe purpose was to compare two approaches for the acquisition and analysis of dynamic-contrast-enhanced MRI data with respect to differences in the modelling of the arterial input-function (AIF), the dependency of the model parameters on physiological parameters and their numerical stability. Eight hundred tissue concentration curves were simulated for different combinations of perfusion, permeability, interstitial volume and plasma volume based on two measured AIFs and analysed according to the two commonly used approaches. The transfer constants (Approach 1) K (trans) and (Approach 2) k (ep) were correlated with all tissue parameters. K (trans) showed a stronger dependency on perfusion, and k (ep) on permeability. The volume parameters (Approach 1) v (e) and (Approach 2) A were mainly influenced by the interstitial and plasma volume. Both approaches allow only rough characterisation of tissue microcirculation and microvasculature. Approach 2 seems to be somewhat more robust than 1, mainly due to the different methods of CA administration.

• Tumor perfusion assessed by dynamic contrast-enhanced MRI correlates to the grading of renal cell carcinoma: Initial results

  Authors: M Palmowski, I. Schifferdecker, S. Zwick, S Macher-Göppinger, H. Laue, A Haferkamp, H U Kauczor, F Kiessling, P Hallscheidt

  Fraunhofer MEVIS.

  In this study, we investigated whether assessment of the tumor perfusion by dynamic contrast-enhanced magnetic resonance imaging (DCE MRI) enables to estimate the morphologic grading of renal cellIn this study, we investigated whether assessment of the tumor perfusion by dynamic contrast-enhanced magnetic resonance imaging (DCE MRI) enables to estimate the morphologic grading of renal cell carcinomas. A total of 21 patients with suspected renal cell cancer were examined using a Gadobutrol-enhanced, dynamic saturation-recovery, turbo-fast, low-angle shot sequence. Tumor perfusion and the tissue-blood ratio within the entire tumor and the most highly vascularized part of the tumor were calculated according to the model of Miles. Immediately after examination, patients underwent surgery, and the results from imaging were compared with the morphological analysis of the histologic grading. Fourteen patients had G2 tumors, and seven patients had G3 tumors. Significantly higher perfusion values (p < 0.05) were obtained in G3 tumors than in G2 tumors when the entire tumor area was considered (1.59 ± 0.44 (ml/g/min) vs. 1.08 ± 0.38 (ml/g/min)) or its most highly vascularized part (2.14 ± 0.89 (ml/g/min) vs. 1.40 ± 0.49 (ml/g/min)). By contrast, the tissue-blood ratios did not differ significantly between the two groups. In conclusion, unlike tissue-blood ratio, surrogate parameters of the tumor perfusion determined by DCE MRI seem to allow an estimation of the grading of renal cell carcinoma. However, further studies with high case numbers and including patients with G1 tumors are required to evaluate the full potential and clinical impact.