@inproceedings{BourierSchneiderHeissenhuberetal., author = {Bourier, F. and Schneider, H.-J. and Heißenhuber, F. and Ganslmeier, P. and Brost, A. and Koch, M. and Hornegger, J. and Kleinoeder, A. and Kiraly, A. and Barbot, J. and Strobel, Norbert and Kurzidim, K.}, title = {Augmented Fluoroscopy to Guide Transseptal Puncture}, series = {Venice Arrhythmias 2011 01/2011}, booktitle = {Venice Arrhythmias 2011 01/2011}, language = {en} } @inproceedings{BourierSchneiderHeissenhuberetal., author = {Bourier, F. and Schneider, H.-J. and Heißenhuber, F. and Ganslmeier, P. and Brost, A. and Koch, M. and Hornegger, J. and Kleinoeder, A. and Kiraly, A. and Barbot, J. and Strobel, Norbert and Kurzidim, K.}, title = {Pulmonary Vein Isolation guided by fluoroscopybased 3D-Navigation on a biplane angiography system}, series = {Venice Arrhythmias 2011 01/2011}, booktitle = {Venice Arrhythmias 2011 01/2011}, language = {en} } @inproceedings{BrostWimmerLiaoetal., author = {Brost, A. and Wimmer, A. and Liao, R. and Hornegger, J. and Strobel, Norbert}, title = {Catheter Tracking: Filter-Based vs. Learning-Based}, series = {M. Goesele, S. Roth, A. Kuijper, B. Schiele, K. Schindler (eds.) Pattern Recognintion, LNCS, vol. 6376}, booktitle = {M. Goesele, S. Roth, A. Kuijper, B. Schiele, K. Schindler (eds.) Pattern Recognintion, LNCS, vol. 6376}, publisher = {Springer}, address = {Berlin, Heidelberg}, isbn = {978-3-642-15985-5}, doi = {https://doi.org/10.1007/978-3-642-15986-2_30}, pages = {293 -- 302}, language = {en} } @incollection{HoffmannMuellerKurzidimetal., author = {Hoffmann, Matthias and M{\"u}ller, Simone and Kurzidim, Klaus and Strobel, Norbert and Hornegger, Joachim}, title = {Robust Identification of Contrasted Frames in Fluoroscopic Images}, series = {Bildverarbeitung f{\"u}r die Medizin 2015}, booktitle = {Bildverarbeitung f{\"u}r die Medizin 2015}, publisher = {Springer Vieweg}, address = {Berlin, Heidelberg}, isbn = {978-3-662-46223-2}, doi = {10.1007/978-3-662-46224-9_6}, abstract = {For automatic registration of 3-D models of the left atrium to fluoroscopic images, a reliable classification of images containing contrast agent is necessary. Inspired by previous approaches on contrast agent detection, we propose a learning-based framework which is able to classify contrasted frames more robustly than previous methods, Furthermore, we performed a quantitative evaluation on a clinical data set consisting of 34 angiographies. Our learning-based approach reached a classification rate of 79.5\%. The beginning of a contrast injection was detected correctly in 79.4\%.}, language = {en} } @inproceedings{HohlBoeseStrobeletal., author = {Hohl, C. and Boese, Jan and Strobel, Norbert and Banckwitz, R. and Lauritsch, G. and M{\"u}hlenbruch, G. and G{\"u}nther, R. W.}, title = {Angiographische CT: Messung der Patientendosis}, series = {R{\"o}Fo: Fortschritte auf dem Gebiet R{\"o}ntgenstrahlen und bildgebenden Verfahren}, volume = {180}, booktitle = {R{\"o}Fo: Fortschritte auf dem Gebiet R{\"o}ntgenstrahlen und bildgebenden Verfahren}, number = {S 1}, doi = {10.1055/s-2008-1073628}, abstract = {Ziele: Messung der Organ- und Effektivdosen bei Angiographischen CT-Untersuchungen (ACT) des Kopfs, des Herzens, der Leber sowie der Lendenwirbels{\"a}ule (LWS) mit einem anthropomorphen Phantom und Thermolumineszenz Dosimetern (TLD). Methode: ACT erlaubt die Schnittbildgebung von Weichteilstrukturen bei der Verwendung einer C-Bogen Durchleuchtungseinheit mit Flachbilddetektor. Um die sowohl die Organdosen als auch die effektiven Dosen bei ACT-Untersuchungen des Kopfs, des Herzens, der Leber und der LWS zu bestimmen, wurde ein m{\"a}nnliches Alderson-Rando-Phantom mit jeweils 129 TLD an 43 Messstellen best{\"u}ckt. Die Messungen wurden an zwei ACT-Systemen (AXIOM Artis, VB31C, Siemens, Forchheim) mit verschiedenen Detektorgr{\"o}ßen (20×20cm und 30×40cm) durchgef{\"u}hrt. Mit dem 20×20 Detektor wurden der Kopf (70kV, 126 frames, 0,36µGy/frame, 7,9ms/frame), die LWS (90kV, 275 frames, 0,36Gy/frame, 4,9ms/frame) und das Herz (70kV, 126 frames, 0,54µGy/frame, 10,9ms/frame) untersucht. Mit dem 30×40 Detektor wurden der Kopf (84kV, 496 frames, 1,2µGy/frame, 11,6ms/frame) die Leber (90kV, 397 frames, 0,36µGy/frame, 4,9ms/frame) und das Herz (70kV, 248 frames, 0,54µGy/frame, 4,9ms/frame) untersucht. Zus{\"a}tzlich wurde mit dem 30×40 Detektor noch eine EKG-getriggerte Herz-Untersuchung durchgef{\"u}hrt (992 frames, Phasen-Zentrum 75\%, Phasen-L{\"a}nge 50\%, 50\% mAs-Modulation). Nach jeder Messung wurden die TLD gewechselt und ausgewertet. Gem{\"a}ß den ICRP-Empfehlungen sind Organ- und Effektivdosen berechnet worden, die dann mit bekannten Dosiswerten vergleichbarer MSCT-Untersuchungen korreliert wurden. Ergebnis: Die Messungen der effektiven Dosis f{\"u}r den 20×20 Detektor ergaben Werte von 0,2mSv (Kopf), 2,76mSv (LWS) und 1,25mSv (Herz). Mit dem 30×40 Detektor ergaben sich Werte von 2,95mSv (Kopf), 6,71mSv (Leber) und 3,31mSv (Herz, 248 frames). F{\"u}r die EKG-getriggerte Untersuchung des Herzens ergab sich eine effektive Dosis von 10,88 mSv bei 4 Uml{\"a}ufen (992 frames). Vergleichbare MSCT-Untersuchung belaufen sich auf 2,9mSv (Kopf), 4,8mSv (LWS), 4,3mSv (Leber) und 8,3mSv (Herz). Schlussfolgerung: Da ACT-Systeme mit einer Belichtungsautomatik ausgestattet sind, h{\"a}ngt die effektive Dosis stark vom Patientendurchmesser ab. Bei Messungen mit dem anthropomorphen Phantom konnte gezeigt werden, dass die effektiven Dosen von ACT-Untersuchungen sogar noch unterhalb von vergleichbaren MSCT-Untersuchungen liegen k{\"o}nnen. Allerdings muss dabei ber{\"u}cksichtigt werden, dass die Bildqualit{\"a}t und insbesondere der Weichteilkontrast nicht vergleichbar sind.}, language = {de} } @article{BourierFahrigWangetal., author = {Bourier, Felix and Fahrig, Rebecca and Wang, Paul and Santangeli, Pasquale and Kurzidim, Klaus and Strobel, Norbert and Moore, Teri and Hinkel, Cameron and Al-Ahmad, Amin}, title = {Accuracy assessment of catheter guidance technology in electrophysiology procedures: a comparison of a new 3D-based fluoroscopy navigation system to current electroanatomic mapping system}, series = {Journal of Cardiovascular Electrophysiology}, volume = {25}, journal = {Journal of Cardiovascular Electrophysiology}, number = {1}, doi = {10.1111/jce.12264}, pages = {74 -- 83}, abstract = {Background: With increasing complexity in electrophysiology (EP) procedures, the use of electroanatomic mapping systems (EAMS) as a supplement to fluoroscopy has become common practice. This is the first study that evaluates spatial and point localization accuracy for 2 current EAMS, CARTO3(®) (Biosense Webster, Diamond Bar, CA, USA) and EnSite Velocity(®) (St. Jude Medical Inc., St. Paul, MN, USA), and for a novel overlay guidance (OG) software (Siemens AG, Forchheim, Germany) in a phantom experiment. Methods and results: A C-arm CT scan was performed on an acrylic phantom containing holes and location markers. Spatial accuracy was assessed for each system using distance measurements involving known markers inside the phantom and properly placed catheters. Anatomical maps of the phantom were acquired by each EAMS, whereas the 3D-based OG software superimposed an overlay image of the phantom, segmented from the C-arm CT data set, onto biplane fluoroscopy. Registration processes and landmark measurements quantitatively assessed the spatial accuracy of each technology with respect to the ground truth phantom. Point localization performance was 0.49 ± 0.25 mm in OG, 0.46 ± 0.17 mm in CARTO3(®) and 0.79 ± 0.83 mm in EnSite(®) . The registration offset between virtual visualization and reality was 1.10 ± 0.52 mm in OG, 1.62 ± 0.77 mm in CARTO3(®) and 2.02 ± 1.21 mm in EnSite(®) . The offset to phantom C-arm CT landmark measurements was 0.30 ± 0.26 mm in OG, 0.24 ± 0.21 mm in CARTO3(®) and 1.32 ± 0.98 mm in EnSite(®) . Conclusions: Each of the evaluated EP guidance systems showed a high level of accuracy; the observed offsets between the virtual 3D visualization and the real phantom were below a clinically relevant threshold of 3 mm.}, language = {en} } @article{MeyerBrostKraitchmanetal., author = {Meyer, Bernhard Christian and Brost, Alexander and Kraitchman, Dara L and Gilson, Wesley D and Strobel, Norbert and Hornegger, Joachim and Lewin, Jonathan S and Wacker, Frank K}, title = {Percutaneous punctures with MR imaging guidance: comparison between MR imaging-enhanced fluoroscopic guidance and real-time MR Imaging guidance}, series = {Radiology}, volume = {266}, journal = {Radiology}, number = {3}, doi = {10.1148/radiol.12120117}, pages = {912 -- 919}, abstract = {Purpose: To evaluate and compare the technical accuracy and feasibility of magnetic resonance (MR) imaging-enhanced fluoroscopic guidance and real-time MR imaging guidance for percutaneous puncture procedures in phantoms and animals. Materials and methods: The experimental protocol was approved by the institutional animal care and use committee. Punctures were performed in phantoms, aiming for markers (20 each for MR imaging-enhanced fluoroscopic guidance and real-time MR imaging guidance), and pigs, aiming for anatomic landmarks (10 for MR imaging-enhanced fluoroscopic guidance and five for MR imaging guidance). To guide the punctures, T1-weighted three-dimensional (3D) MR images of the phantom or pig were acquired. Additional axial and coronal T2-weighted images were used to visualize the anatomy in the animals. For MR imaging-enhanced fluoroscopic guidance, phantoms and pigs were transferred to the fluoroscopic system after initial MR imaging and C-arm computed tomography (CT) was performed. C-arm CT and MR imaging data sets were coregistered. Prototype navigation software was used to plan a puncture path with use of MR images and to superimpose it on fluoroscopic images. For real-time MR imaging, an interventional MR imaging prototype for interactive real-time section position navigation was used. Punctures were performed within the magnet bore. After completion, 3D MR imaging was performed to evaluate the accuracy of insertions. Puncture durations were compared by using the log-rank test. The Mann-Whitney U test was applied to compare the spatial errors. Results: In phantoms, the mean total error was 8.6 mm ± 2.8 with MR imaging-enhanced fluoroscopic guidance and 4.0 mm ± 1.2 with real-time MR imaging guidance (P < .001). The mean puncture time was 2 minutes 10 seconds ± 44 seconds with MR imaging-enhanced fluoroscopic guidance and 37 seconds ± 14 with real-time MR imaging guidance (P < .001). In the animal study, a tolerable distance (<1 cm) between target and needle tip was observed for both MR imaging-enhanced fluoroscopic guidance and real-time MR imaging guidance. The mean total error was 7.7 mm ± 2.4 with MR imaging-enhanced fluoroscopic guidance and 7.9 mm ± 4.9 with real-time MR imaging guidance (P = .77). The mean puncture time was 5 minutes 43 seconds ± 2 minutes 7 seconds with MR imaging-enhanced fluoroscopic guidance and 5 minutes 14 seconds ± 2 minutes 25 seconds with real-time MR imaging guidance (P = .68). Conclusion: Both MR imaging-enhanced fluoroscopic guidance and real-time MR imaging guidance demonstrated reasonable and similar accuracy in guiding needle placement to selected targets in phantoms and animals.}, language = {en} } @article{BrostRaabKleinoederetal., author = {Brost, Alexander and Raab, Jens and Kleinoeder, Andreas and Kurzendorfer, Tanja and Bourier, Felix and Koch, Martin and Hoffmann, Matthias and Strobel, Norbert and Kurzidim, Klaus and Hornegger, Joachim}, title = {Medizinische Bildverarbeitung f{\"u}r die minimal-invasive Behandlung von Vorhofflimmern}, series = {Deutsche Zeitschrift f{\"u}r klinische Forschung, Innovation und Praxis (DZKF)}, volume = {17}, journal = {Deutsche Zeitschrift f{\"u}r klinische Forschung, Innovation und Praxis (DZKF)}, number = {6}, publisher = {Mediengruppe Oberfranken Fachverlage GmbH \& Co. KG}, pages = {36 -- 41}, language = {en} } @article{BourierVukajlovicBrostetal., author = {Bourier, Felix and Vukajlovic, Dejan and Brost, Alexander and Hornegger, Joachim and Strobel, Norbert and Kurzidim, Klaus}, title = {Pulmonary vein isolation supported by MRI-derived 3D-augmented biplane fluoroscopy: a feasibility study and a quantitative analysis of the accuracy of the technique}, series = {Journal of Cardiovascular Electrophysiology}, volume = {24}, journal = {Journal of Cardiovascular Electrophysiology}, number = {2}, doi = {10.1111/jce.12009}, pages = {113 -- 120}, abstract = {Background: Despite the advancement of technology in electroanatomic mapping systems (EAMS), fluoroscopy remains a necessary, basic imaging modality for electrophysiology procedures. We present a feasibility study of new software that enables 3D-augmented fluoroscopy in biplane catheterization laboratories for planning and guidance of pulmonary vein isolation (PVI). The computer-assisted overlay registration accuracy was assessed in a clinical setting using an automatic calculation of overlay projection geometry that was derived from hardware sensors in C-arms, detectors, and patient table. Methods: Consecutive patients (n = 89) underwent left atrium (LA) magnetic resonance imaging MRI scan prior to PVI. Ideal ablation lines encircling the ipsilateral pulmonary veins (PVs) at antral level were drawn onto the segmented LA surface. The 3D-model was superimposed onto biplane fluoroscopy and matched with angiographies of LA and PVs. Three-dimensional-overlay projection geometry was automatically calculated from C-arm, detectors, and table sensors. Accuracy of technique was assessed as alignment of MRI-derived 3D overlay and angiographic LA/PV anatomy. Integrity of registered overlay was quantified using landmark measurements. Results: Alignment offsets were 1.3 ± 1.5 mm in left PV, 1.2 ± 1.5 mm in right PV, and 1.1 ± 1.4 mm in LA roof region. Bravais-Pearson correlation of the landmark measurements was r = 0.978 (s < 0.01), mean offset between landmark distance measurements was 1.4 ± 0.78 mm. Average time needed for overlay registration was 9.5 ± 3.5 seconds. Conclusions: MRI-derived 3D-augmented fluoroscopy demonstrated a high level of accuracy when compared with LA/PV angiography. The new system could be especially useful to guide procedures not supported by EAMS, such as cryotechnique PVI.}, language = {en} } @article{BrostWimmerLiaoetal., author = {Brost, Joachim and Wimmer, Andreas and Liao, Rui and Bourier, Felix and Koch, Martin and Strobel, Norbert and Kurzidim, Klaus and Hornegger, Joachim}, title = {Constrained registration for motion compensation in atrial fibrillation ablation procedures}, series = {IEEE Trans Med Imaging}, volume = {31}, journal = {IEEE Trans Med Imaging}, number = {4}, doi = {10.1109/tmi.2011.2181184}, pages = {870 -- 881}, abstract = {Fluoroscopic overlay images rendered from preoperative volumetric data can provide additional anatomical details to guide physicians during catheter ablation procedures for treatment of atrial fibrillation (AFib). As these overlay images are often compromised by cardiac and respiratory motion, motion compensation methods are needed to keep the overlay images in sync with the fluoroscopic images. So far, these approaches have either required simultaneous biplane imaging for 3-D motion compensation, or in case of monoplane X-ray imaging, provided only a limited 2-D functionality. To overcome the downsides of the previously suggested methods, we propose an approach that facilitates a full 3-D motion compensation even if only monoplane X-ray images are available. To this end, we use a training phase that employs a biplane sequence to establish a patient specific motion model. Afterwards, a constrained model-based 2-D/3-D registration method is used to track a circumferential mapping catheter. This device is commonly used for AFib catheter ablation procedures. Based on the experiments on real patient data, we found that our constrained monoplane 2-D/3-D registration outperformed the unconstrained counterpart and yielded an average 2-D tracking error of 0.6 mm and an average 3-D tracking error of 1.6 mm. The unconstrained 2-D/3-D registration technique yielded a similar 2-D performance, but the 3-D tracking error increased to 3.2 mm mostly due to wrongly estimated 3-D motion components in X-ray view direction. Compared to the conventional 2-D monoplane method, the proposed method provides a more seamless workflow by removing the need for catheter model re-initialization otherwise required when the C-arm view orientation changes. In addition, the proposed method can be straightforwardly combined with the previously introduced biplane motion compensation technique to obtain a good trade-off between accuracy and radiation dose reduction.}, language = {en} }