AG Artificial Intelligence in Cardiovascular Medicine

We have the ambitious goal to radically improve surgical training.
In order to achieve that, we enhance surgical training with augmented reality concepts that make surgical training more realistic (we coined this approach ‚Hyperrealism‘) and provide quantitative information about the training process itself.
Another of our core research topics is to build patient-specific simulators for surgical training to make training more effective and procedures safer for the patient.

Supplemental Material:

• tempCycleGAN: Generative Adversarial Network Improves Realism During Surgical Training - MICCAI 2018 (Youtube)
• Training Reconstructive Mitral Valve Surgery on Patient-Specific Mitral Valve Models (Youtube)

Heart Valve Modelling and Visualization, Computer-assistance in Heart Valve Surgery

Heart valves are complex and highly dynamical anatomical structures. We are developing methods for extracting such complicated geometrical information from 3D+t echocardiographic data. Furthermore, our aim is to quantitatively and qualitatively present the patient information to provide deeper insights into pathological changes of individual valves.

Beyond that, we are working on computer-based intraoperative assistance modules for mitral valve repair to enable decision support e.g. for optimal prosthesis selection.

Supplemental Material:
 

• Temporal Views of Flattened Mitral Valve Geometries (Youtube)
• Video " From intraoperative empirical valve inspection towards quantitative optical tracking analysis
• Geförderte Projekte der DFG

Magnetic resonance imaging (MRI) is a valuable tool to non-invasively assess geometrical, contraction-related, and tissue-related pathological changes.

By the help of machine learning techniques, we characterize these properties to better predict therapy response.

Invasive heart catheters are a common approach to diagnose coronary artery disease. However, 3D-computed tomography (CT) can provide equal information on stenotic regions and plaque by being less invasive at the same time. Our aim is to analyse these data sets by the help of radiomics and deep learning methods.

Echocardiography is the most employed modality in cardiovascular diseases due to its excellent capabilities of resolving anatomy and motion in real-time at all stages of the treatment process. Unfortunately, the analysis of the data is often hampered by image artefacts and a small field of view. Within the research campus STIMULATE, we work on deep learning-based image compounding and registration approaches to overcome these issues.
 

• www.forschungscampus-stimulate.de

Lalith Sharan

Robert Kreher (extern Research Campus STIMULATE Magdeburg)

Sven Koehler

Roger Fredrick Karl

Samantha Fischer (Medicine)

Simon Sauerzapf

Jonathan Kloss

Ulrike Schnaithmann

Antonia Stern

Florian Ritzmann

Jean-Luc Busch

Lukas Burger

Alexander Rogausch

Koehler, S., Tandon, A., Hussain, T., Latus, H., Pickardt T., Sarikouch, S., Beerbaum, B., Greil, G., Engelhardt, S., and Wolf, I. How well do U-Net-based segmentation trained on adult cardiac magnetic resonance imaging data generalize to rare congenital heart diseases for surgical planning? Proc. SPIE 11315, Medical Imaging 2020: Image-Guided Procedures, Robotic Interventions, and Modeling, 113151K (16 March 2020) doi.org/10.1117/12.2550651

Preetha, C.J., Wehrtmann, F.S., Sharan, L., Fan, C., Kloss, J., Müller-Stich, B.P., Nickel, F., Engelhardt, S., Towards augmented reality-based suturing in monocular laparoscopic training. Proc. SPIE 11315, Medical Imaging 2020: Image-Guided Procedures, Robotic Interventions, and Modeling, 113150X (16 March 2020); doi.org/10.1117/12.2550830

Eulzer, P., Engelhardt, S., Lichtenberg, N., De Simone, R., Lawonn, K., Temporal Views of Flattened Mitral Valve Geometries, IEEE Trans Vis Comput Graph. 2020 Jan; 26(1):971-980. doi: 10.1109/TVCG.2019.2934337.

Engelhardt, S., Sauerzapf, S., Preim, B., Karck, M., Wolf, I., De Simone, R., Flexible and Comprehensive Patient-Specific Mitral Valve Silicone Models with Chordae Tendinae Made From 3D-Printable Molds. IJCARS Special Issue IPCAI 2019. 14(7):1177–1186, doi: 10.1007/s11548-019-01971-9

Bernard, O., Lalande, A., Zotti, C., Cervenansky, F., Yang, X., Heng, P.A., Cetin, I., Lekadir, K., Camara, O., Gonzalez Ballester, M.A.; Sanroma, G., Napel, S., Petersen, S., Tziritas, G., Grinias, E., Khened, M., Kollerathu, V.A., Krishnamurthi, G., Rohé, M.M., Pennec, X; Sermesant, M., Isensee, F., Jäger, P., Maier-Hein K.H., Full, P.M., Wolf, I., Engelhardt, S.,  Baumgartner, C.F., Koch, L.M., Wolterink, J.M., Išgum, I., Jang, Y., Hong, Y., Patravali, J., Jain, S., Humbert, O., Jodoin, P-M., Deep Learning Techniques for Automatic MRI Cardiac Multi-structures Segmentation and Diagnosis: Is the Problem Solved?, IEEE Transactions on Medical Imaging  37:11 , 2018; 2514 - 2525 Doi: 10.1109/TMI.2018.2837502

Engelhardt, S., Sharan, L., Karck, M., De Simone, R., Wolf, I., Cross-Domain Conditional Generative Adversarial Networks for Stereoscopic Hyperrealism in Surgical Training. In: Shen D. et al. (eds) Medical Image Computing and Computer Assisted Intervention – MICCAI 2019. MICCAI 2019. Lecture Notes in Computer Science, vol 11768. Springer, Cham, pp 155-163, doi: doi.org/10.1007/978-3-030-32254-0_18

Engelhardt S., De Simone R., Full P.M., Karck M., Wolf I. (2018) Improving Surgical Training Phantoms by Hyperrealism: Deep Unpaired Image-to-Image Translation from Real Surgeries. In: Frangi A., Schnabel J., Davatzikos C., Alberola-López C., Fichtinger G. (eds) Medical Image Computing and Computer Assisted Intervention – MICCAI 2018. MICCAI 2018. Lecture Notes in Computer Science, vol 11070. Springer, Cham, doi: 10.1007/978-3-030-00928-1_84

• on google scholar