MIDL 2021

International Conference on
Medical Imaging with Deep Learning
5. bis 7. Juli 2021

Deutsche Forschungsgemeinschaft

Development and establishment of the National Metadata Repositories (NMDR)

Clinical research is increasingly dependent on consistently defined characteristics, e.g. for the re-use of clinical data from the supply in studies, the pooling of study data in multicenter studies or even the implementation of meta-studies. Internationally, there is a considerable effort to establish metadata repositories (MDR) or official registers in order to designate relevant data elements such as a blood pressure uniformly, incl. definitions and annotations with codes of standardized vocabularies. Based on this, software services can be developed to provide semantic data integration.

This challenge was also recognized by the Technology and Methods Platform for Networked Medical Research (TMF), which has already funded preliminary work on this subject. The strategic goal of this DFG project is to establish a collaborative, quality-assured, neutral, permanent, free and accessible metadata register for clinical and epidemiological research in Germany. The categories mentioned are the result of a requirement analysis, which was conducted as part of the TMF project "Community-evaluation MDR". The envisioned results should support all clinical researchers, science-initiated studies, registers or cohorts that are dependent on high-quality data . At the same time, these points are not sufficiently addressed by any existing system.

Participants include scientists from Leipzig, Lübeck, Jena, Heidelberg and Berlin.

Selected Publications:

  1. Ulrich H., Kock A.K., Duhm-Harbeck P., Habermann J.K., Ingenerf J.
    Metadata Repository for Improved Data Sharing and Reuse Based on HL7 FHIR
    In: MIE-Conference (scoped by HEC 2016) in Munich [paper accepted]; 2016


M.Sc. A.-K. Kock-Schoppenhauer
M.Sc. H. Ulrich
Prof. Dr. J. Ingenerf


Created at March 16, 2017 - 3:20pm.

Towards Realtime MR-guided Motion Compensation using Model-based Registration without Fiducial Markers

Physiological patient motion is an important problem in accurate dose delivering during radiotherapy. Accurate and realtime motion compensation based on image-guidance could be realised in a combined MR-radiotherapy treatment setup. The objective of this project is to develop algorithms that can estimate intra-fraction motion reliably without implanted fiducial markers and improve on state-of-the-art techniques in terms of accuracy and computational speed.

In contrast to previous work, which predominantly used template matching to achieve realtime speed, we propose to incorporate prior knowledge as well as patient-specific information of plausible deformations during motion estimation. Superior motion estimation, especially for peripheral organs of risk, will be achieved using these models. To account for motion variability, MR images with high temporal resolution acquired during a short setup phase under free breathing could be incorporated for a patient-specific training. Building upon previous work, a motion-model based on principal component analysis or a Bayesian framework can be robustly trained using highly efficient deformable registration. Multiple distributed keypoints at discriminative geometric locations will be extracted automatically using machine learning techniques to avoid the need for invasive implantation of fiducial markers. Robust and accurate realtime motion estimation will be performed within a computationally efficient optimisation framework that incorporates the training model for plausible regularised motion estimation and avoids tracking errors by sampling a large space of potential motion vectors. The algorithms will be validated on retrospective clinical 4D MRI scans using manually annotated landmarks to demonstrate its suitability and advances over state-of-the-art methods.

Fig. 1: Overview of project plan for motion estimation in 4D-MRI using model-based regularisation

The project is funded by Deutsche Forschungsgemeinschaft (DFG) (HE 7364/1-1).

Selected Publications:

  1. Wilms M., Ha I.Y., Handels H., Heinrich M.P.
    Model-based Regularisation for Respiratory Motion Estimation with Sparse Features in Image-guided Interventions
    MICCAI 2016

  2. Ha I.Y., Wilms M., Heinrich M.P.
    Multi-object segmentation in chest X-ray using cascaded regression ferns
    BVM 2017

  3. Heinrich M.P., Papiez B.W., Schnabel J., Handels H.
    Non-Parametric Discrete Registration with Convex Optimisation
    WBIR 2014

Project Team:

M.Sc. In Young Ha
M.Sc. Matthias Wilms
Jun.-Prof. Dr. Mattias P. Heinrich


Created at January 23, 2017 - 11:40am by Heinrich. Last modified at February 23, 2017 - 4:53pm.

Learning Contrast-invariant Contextual Local Descriptors and Similarity Metrics for Multi-modal Image Registration

Deformable image registration is a key component for clinical imaging applications involving multi-modal image fusion, estimation of local deformations and image-guided interventions. A particular challenge for establishing correspondences between scans from different modalities: magnetic resonance imaging (MRI), computer tomography (CT) or ultrasound, is the definition of image similarity. Relying directly on intensity differences is not sufficient for most clinical images, which exhibit non-uniform changes in contrast, image noise, intensity distortions, artefacts, and globally non-linear intensity relations (for different modalities).

In this project algorithms with increased robustness for medical image registration will be developed. We will improve on current state-of-the-art similarity measures by combining a larger number of versatile image features using simple local patch or histogram distances. Contrast-invariance and strong discrimination between corresponding and non-matching regions will be reached by capturing contextual information through pair-wise comparisons within an extended spatial neighbourhood of each voxel. Recent advances in machine learning will be used to learn problem-specific binary descriptors in a semi-supervised manner that can improve upon hand-crafted features by including a priori knowledge. Metric learning and higher-order mutual information will be employed for finding mappings between feature vectors across scans in order to reveal new relations among feature dimensions. Employing binary descriptors and sparse feature selection will improve computational efficiency (because it enables the use of the Hamming distance), while maintaining the robustness of the proposed methods.

A deeper understanding of models for image similarity will be reached during the course of this project. The development of new methods for currently challenging (multi-modal) medical image registration problems will open new perspectives of computer-aided applications in clinical practice, including multi-modal diagnosis, modality synthesis, and image-guided interventions or radiotherapy.

Abb. 1: Overview of project plan for learning multi-modal metrics using correspondences in aligned training data.

The project is funded by Deutsche Forschungsgemeinschaft (DFG) (HE 7364/2-1).

Selected Publications:

  1. Heinrich M.P., Blendowski M.
    Multi-Organ Segmentation using Vantage Point Forests and Binary Context Features
    MICCAI 2016

  2. Blendowski M., Heinrich M.P.
    Kombination binärer Kontextfeatures mit Vantage Point Forests zur Multi-Organ-Segmentierung
    BVM 2017

  3. Heinrich M.P., Jenkinson M., Bhushan M., Matin T., Gleeson F.V., Brady S.M., Schnabel J.A.
    MIND: modality independent neighbourhood descriptor for multi-modal deformable registration.
    Medical image analysis 2012

  4. Heinrich M.P., Jenkinson M., Papiez B.W., Brady S.M., Schnabel J.A.
    Towards realtime multimodal fusion for image-guided interventions using self-similarities
    MICCAI 2013

Project Team:

M.Sc. Max Blendowski
Jun.-Prof. Dr. Mattias P. Heinrich


Created at January 23, 2017 - 11:40am by Heinrich. Last modified at February 23, 2017 - 4:56pm.

Integrated Analysis and Probabilistic Registration of Medical Images with Missing Correspondences

The automatic, robust and reliable registration of medical images is a central problem in medical image computing with high impact on image-guided diagnostics and therapy. Currently available registration methods reach their limits, if strong anatomical or pathologic discrepancies are present in the images and corresponding structures are missing in parts of the images. Another limitation of current registration methods is the lack of information they provide to the user about the local (un)certainty of the estimated transformation and therefore does not allow an assessment of the registration results.

The aim of this project is to enable the robust and reliable registration of images even if one-to-one correspondences are missing in parts of the images. To achieve this, a general probabilistic registration framework based on correspondence probabilities is developed that does not only rely on image intensities but also on additional information extracted by image analysis methods like organ segmentations, landmarks and local image features to align images. The methods to develop will enable the registration of areas with missing local correspondences as well as the objective assessment of the reliability of the local registration results.

The proposed methodical innovations extend the medical application spectrum of image registration algorithms, significantly. For example, the proposed method will facilitate and improve the quality of image-based follow-up studies and clinical monitoring, comparison of pre- and post-operative images as well as image-based statistical studies to reveal spatial distribution patterns of pathological tissues or neuronal activities.

Project Team:

M.Sc. Sandra Schultz
Dr. rer. nat. Jan Ehrhardt
Prof. Dr. rer. nat. habil. Heinz Handels


Created at April 26, 2016 - 10:34am. Last modified at May 23, 2016 - 3:39pm.

Patient-Individual 4D Virtual Reality Simulation of Punctures and Radio-Frequency Ablations in Virtual Body Models with Breating Motion

Within the project we research methods for realistic 4D visuo-haptic VR simulation Virtual body models are animated by respiratory motion models. Applications are the patient-individual planning and training of punctures and radio-frequency ablation under respiratory motion. Static 3D image data sets of the patient are animated both by individual as well as mean-4D motion models that have been extracted from 4D image data and serve as a voxel-based description of real breathing movements. Using surrogate based 4D motion models regards also the variability of breathing in different respiratory cycles. By means of non-linear registration methods, the anatomical differences between the model and the patient's anatomy are compensated and the motion fields are transferred to animate the static 3D patient data. The 4D-motion models are integrated in a visuo-haptic framework the haptic-visual-driven interaction allows the puncture and ablation needle to interact with the breathing virtual body. For visuo-haptic 4D representation of moving 3D image data in real time special volume-based 4D rendering techniques are developed and optimized for the GPU. Furthermore near the diaphragm region, the effects of the respiratory movement on the biophysical simulation of RF ablations and the 4D needle path planning are compared to the planning and simulation in static 3D data. In addition to the evaluation of the different methods and system components, a user study about the VR training simulator enhanced by breathing motion is finally conducted.

Fig. 1: 4D image sequences showing needle and breathing motion.

The project is funded by the German Research Foundation (DFG: HA 2355/11-2).

Project team:

Dr. Andre Mastmeyer
Prof. Dr. Heinz Handels


Created at June 12, 2015 - 4:00pm by Mastmeyer.

Segmentierung von Hirngefäßen und Blutflussanalyse in der 4D-Magnetresonanzangiographie bei zerebralen arteriovenösen Malformationen - Untersuchungen zu Hämodynamik und Gewebemarkern

Zur Planung einer invasiven Therapie für Patienten mit Gefäßfehlbildungen des Gehirns, sog. arteriovenöse Malformationen (Abk.: AVM), ist die Abschätzung des individuellen natürlichen Blutungsrisikos von entscheidender Bedeutung. Im Rahmen des Projektes wurden neue Methoden zur Segmentierung von AVMs in 3D-TOF-MRA-Bilddaten sowie zur Analyse des Blutflsuses in 4D-TREAT-MR-Bilddaten entwickelt und zur Auswertung im Rahmen von Studien in ein Softwaresystem namens AnToNIa (Abk. f.: Analysis Tool for Neuro Imaging Data) integriert.  Mithilfe der hier verfügbaren Bildanalyse- und Visualisierungsmethoden ist eine Quantifizierung und dreidimensionale Darstellung des Blutflusses bei AVM-Patienten in hoher räumlicher und zeitlicher Auflösung möglich (Abb. 1).

Abb. 1: Dynamische Darstellung des Bluteinflusses (a-i) auf einem hochaufgelösten 3D-Oberflächenmodell des zerebralen Gefäßsystems

Zur genauen Darstellung und Analyse der räumlichen Struktur des Gefäßsystems im Gehirn konnte durch das neue vierstufige Segmentierungsverfahren unter Einbeziehung von Form- und Intensitätsinformationen eine deutliche Verbesserung gegenüber etablierten Verfahren erreicht werden (Abb. 2). Für die zeitaufgelöste Magnetresonanzangiographie (TWIST/TREAT) wurde ein neues Verfahren der referenzbasierten Kurvenanpassung zur robusten Quantifizierung der Hämodynamik auf Basis von 4D-MRA-Bildsequenzen mit hoher Genauigkeit entwickelt. Im Rahmen einer Monte Carlo Simulation konnte gezeigt werden, dass die Präzision des neuen Verfahrens gegenüber den etablierten Verfahren um 59% gesteigert und dabei die Laufzeit um 33% reduziert werden konnten. Ein weiterer wesentlicher Vorteil des neuen Verfahrens ist die implizite Berücksichtigung der individuellen physiologischen Charakteristika durch die Verwendung einer Referenzkurve.

Abb. 2: 3D-Oberflächenmodell eines zerebralen Gefäßsystems von einem Patienten mit diagnostizierter AVM.

Abb. 3: Farbcodierte Darstellung der extrahierten Werte der Bolus Arrivial Time (BAT) auf einem 3D-Oberflächenmodell (links) und in einer 3D-TOF-MRA-Schicht (rechts). Anhand der BAT-Werte wird erkennbar, welche Gefäße zuerst und welche später durchflossen werden.

Insgesamt wurden innerhalb des Projektes mehr als 50 Patienten mit der TWIST/TREAT untersucht und die Daten mittels der hier der entwickelten Software analysiert. Zunächst wurde der Zusammenhang zwischen den makrovaskulären Fluss und der mikrovaskulären Perfusion um den Nidus herum untersucht. Die Ergebnisse dieser Untersuchung sprechen für zwei Ebenen der Perfusionsbeeinträchtigung: eine makrovaskulär-territoriale und eine mikrovaskulär-lokale Ebene. Darüber hinaus wurde untersucht, ob sich AVMs mit hohem und niedrigem Blutungsrisiko hinsichtlich ihrer hämodynamischen Parameter unterscheiden. Hierbei zeigte sich statistisch robust, dass hohe arterielle Einflussgeschwindigkeiten einen Risikofaktor für eine AVM-Blutung darstellen. Das visuelle Rating und der Vergleich mit der konventionellen Angiographie sind abgeschlossen. Hierbei zeigte sich, dass die dreidimensionale flusskodierte Sichtweise auf die Daten erhebliche Vorteile bietet. Es wurden drei intranidale Flussmuster identifiziert: homogen, uni¬direktional und heterogen.

Die im Rahmen des Forschungsprojektes entwickelten Verfahren und deren Implementierung in ein benutzerfreundliches Auswertetool bilden zudem die Grundlage für diverse weitere Forschungsarbeiten, insbesondere auf dem Gebiet der Hirngefäßaneurysmen.

Das Projekt wird von der Deutschen Forschungsgemeinschaft gefördert (Ha2355/10-1).

Ausgewählte Publikationen

  1. Forkert N.D.,  Illies T., Goebell E., Fiehler J., Säring D., Handels H.,
    Computer-aided Nidus Segmentation and Angiographic Characterization of Arteriovenous Malformations,
    International Journal of Computer Assisted Radiology and Surgery, 8, 775-786, 2013
  2. Forkert N., Schmidt-Richberg A., Fiehler J., Illies T., Möller D., Säring D., Handels H., Ehrhardt J.,
    3D Cerebrovascular Segmentation combining Fuzzy Vessel Enhancement and Level-sets with Anisotropic Energy Weights,
    Magnetic Resonance Imaging, 31, 2, 262-271, 2013
  3. Forkert N., Fiehler J., Illies T., Möller D., Handels H., Säring D.,
    4D Blood Flow Visualization Fusing 3D and 4D MRA Image Sequences,
    Journal of Magnetic Resonance Imaging, 36, 2, 443-53, 2012
  4. Forkert N., Illies T., Möller D., Handels H., Säring D., Fiehler J.,
    Analysis of the Influence of 4D MRA Temporal Resolution on Time-to-Peak Estimation Accuracy for Different Cerebral Vessel Structures,
    American Journal of Neuroradiology, 33(11), 2103-2109, 2012
  5. Forkert N., Fiehler J., Schönfeld M., Sedlacik J., Regelsberger J., Handels H., Illies T.,
    Intranidal Signal Distribution in Post-contrast Time-of-Flight MRA is Associated with Rupture Risk Factors in Arteriovenous Malformations,
    Clinical Neuroradiology, Epub ahead of print, Aug. 2012, Doi 10.1007/s00062-012-0168-8
  6. Forkert N., Kaesemann P., Treszl A., Siemonson S., Cheng B., Handels H., Fiehler J., Thomalla G.,
    Comparison of 10 TTP and Tmax Estimation Techniques for MR Perfusion-Diffusion Mismatch Quantification,
    American Journal of Neuroradiology, 34, 1697-1703, 2012
  7. Forkert N., Schmidt-Richberg A., Fiehler J., Illies T., Möller D., Handels H., Säring D.,
    Automatic Correction of Gaps in Cerebrovascular Segmentations Extracted from 3D Time-of-Flight MRA Datasets,
    Methods of Information in Medicine, 5, 415-422, 2012
  8. Forkert N. Schmidt-Richberg A., Fiehler J., Illies T., Möller D., Handels H., Säring D.,
    Fuzzy-based Vascular Structure Enhancement in Time-of-Flight MRA Images for Improved Cerebrovascular Segmentation,
    Methods of Information in Medicine, 50, 1, 74-83, 2011
  9. Forkert N., Säring D., Handels H.,
    Automatic Analysis of the Anatomy of Arteriovenous Malformations  using 3D and 4D MRA Image Sequences,
    MedInfo 2010, Kapstadt, South Africa, Studies in Health Technology and Informatics, 160, 1268-72, 2010
  10. Forkert N., Säring D., Fiehler J., Illies T., Möller D., Handels H.,
    Automatic Brain Segmentation in Time-of-Flight MRA Images,
    Methods of Information in Medicine, 48, 5, 399-407, 2009
  11. Dennis Säring, Jens Fiehler, Nils Forkert, Merle Piening, Heinz Handels
    Visualization and Analysis of Cerebral Arteriovenous Malformation Combining 3D and 4D MR Image Sequences,
    International Journal of Computer Assisted Radiology and Surgery, 2, 75-79, 2007


Dipl.-Inf. Nils Folkert (Institut für Medizinische Informatik, UKE Hamburg)
Dr. Dennis Säring (Institut für Medizinische Informatik, UKE Hamburg)
Prof. Dr. Heinz Handels


Prof. Dr. med. Jens Fiehler
Dr. med. Till Illies
Klinik für Neuroradiologische Diagnostik und Intervention, UKE


Created at July 16, 2014 - 1:37pm.

Probabilistic Statistical Shape and Appearance Models for Robust Multi-Object Segmentation in Medical Image Data

The objective of the project is to develop a model based method for automatic 3D segmentation of multiple anatomic objects in medical image volumes. The knowledge based segmentation of organs will open up new possibilities in the quantitative radiology, radiation therapy and operation planning. The main focus of this DFG project is the improvement and extension of the statistical shape models with probabilistic point correspondences that had been developed in a previous project by our research group. The probabilistic statistical shape model that already holds the information about the shape of an organ will be extended to a probabilistic shape and appearance model that additionally contains knowledge about the local appearance of organs and the global neighborhood relations between them. Furthermore the probabilistic shape and appearance models will be integrated in an advanced level set segmentation approach to enable a robust and flexible multi-object segmentation of organ ensembles in 3D image volumes.

This projekt is supported by the German Research Foundation (DFG: HA 2355/7-2).


  1. Hufnagel H., Ehrhardt J., Pennec X., Ayache N., Handels H., Coupled Level Set Segmentation Using a Point-Based Statistical Shape Model Relying on Correspondence Probabilities, In: Dawant B. M., Haynor D.R. (eds.), Image Processing, SPIE Medical Imaging 2010, Orlando, Vol. 7623, 1B1-1B8, 2010
  2. Hufnagel H., Ehrhardt J., Pennec X., Ayache N., Handels H.,  Computation of a Probabilistic Statistical Shape Model in a Maximum-a-posteriori Framework, Methods of Information in Medicine,  48, 4, 314-319, 2009
  3. Hufnagel H., Ehrhardt J., Pennec X., Ayache N., Handels H., Level Set Segmentation Using a Point-Based Statistical Shape Model Relying on Correspondence Probabilities, Workshop Probabilistic Models for Medical Image Analysis, PMMIA 09, Medical Image Computing and Computer-Assisted Intervention - MICCAI 2009, London, United Kingdom, 34-44, 2009
  4. Hufnagel H., Pennec X., Ehrhardt J., Ayache N., Handels H., Generation of Statistical Shape Models with Probabilistic Point Correspondences and Expectation Maximization – Iterative Closest Point Algorithm, International Journal of Computer Assisted Radiology and Surgery, 2, 5, 265-273, 2008

Project team:

M.Sc. Julia Krüger
Dr. Jan Ehrhardt
Prof. Dr. Heinz Handels

Created at November 22, 2012 - 3:54pm. Last modified at June 30, 2014 - 11:43am.

4D Medical Image Computing for Image-based Risk Assessment in Radiation Therapy of Moving Tumors

Second phase of the DFG-funded project „4D Medical Image Computing for Model-based Analysis of Respiratory Tumor and Organ Motion“

Respiratory organ and tumor motion is a significant source of error in radiation therapy of the thorax and upper abdomen. In recent years, a variety of technical solutions has been developed to explicitly account for breathing motion during radiation treatment. Methods in clinical use are e. g. so-called gating techniques or respiratory-triggered dose delivery. This means that radiation is only delivered during specified phases of the patients’ breathing cycle. The phases are usually determined using external breathing signals and breathing motion indicators like abdominal bellows or camera-based tracking of surface/skin motion.

External signals, however, are only indicators or surrogates of the inner body tumor motion. Considering especially the risk of intra- and interfractional variations of respiratory motion patterns (and the relationship between tumor motion and breathing signals, respectively) this project aims at investigating the suitability of different motion indicators for predicting tumor motion. Based on 4D CT and 4D MRT image sequences of lung tumor patients acquired over the course of treatment we further intend to quantify dosimetric influences of intra- and interfractional motion variability in standard and gated radiation therapy. Dosimetric consequences in gated radiation therapy are studied simulating the use of different motion indicators – with the final goal of establishing an indicator-specific risk assessment and the development of strategies for combination and optimization of typical breathing motion surrogates (fig. 1).
The project is funded by the German Research Foundation (Deutsche Forschungsgemeinschaft (DFG), HA 2355/9-2).



Fig. 1: Results of a correlation analysis between skin and tumor motion for a lung tumor patient (red: high correlation, green: low correlation). The correlation analysis was based on a 4D CT image se-quence of the patient and the results are used to determine optimal scanline positions of a line laser when applied as a breathing motion indicator in e. g. respiratory-triggered radiation therapy.

Project Team:

M.Sc. Matthias Wilms
Dipl-Inf. Dipl.-Phys. René Werner
Dr. Jan Ehrhardt
Prof. Dr. Heinz Handels

Cooperation Partners:

Prof. Dr. H.-P. Schlemmer, Dr. M. Eichinger / Dr. R. Floca
Abteilung Radiologie / AG Software Development for Integrated Diagnostic and Therapy
Deutsches Krebsforschungszentrum (DKFZ) Heidelberg, Germany

Prof. Dr. Dr. J. Debus, Dr. Dr. C. Thieke
Klinik für Radio-Onkologie und Strahlentherapie
Universitätsklinikum Heidelberg, Germany

Prof. Dr. C. Petersen, Dr. F. Cremers
Klinik und Poliklinik für Strahlentherapie und Radioonkologie
Universitätsklinikum Hamburg-Eppendorf (UKE), Germany



Created at February 20, 2012 - 12:40pm by Kulbe. Last modified at February 21, 2012 - 11:36am.

Patient-specific Virtual Reality Simulation of Punctures Using Puncture Atlases

The aim of this project is to enable the patient-specific virtual reality training of minimal-invasive puncture procedures using special atlases. One topic of the project is the development of multi-altas methods for automated segmentation of the relevant organs and structures. Furthermore, efficient algorithms for volume-based haptic, visual simulation and soft tissue deformation are developed. Soft-tissue deformations should be computed in real-time using a volume-based simulation method. The implementation of computationally expensive algorithms on graphics hardware guarantees the real-time capability of the simulation algorithms.

Within the project a prototype of a VR simulator is set up (Fig. 1). The methods are developed and evaluated in cooperation with clinical partners.

Fig. 1: Immersive VR workstation with shutter glasses and haptic feedback device for puncture training.

The project is funded by the German Research Foundation (DFG: HA 2355/11-1).

Project team:

Prof. Dr. Heinz Handels
M.Sc. Dirk Fortmeier
Dr. Andre Mastmeyer

Created at August 26, 2011 - 4:22pm.

Extended statistical shape models based on probabilistic correspondences for 3D segmentation of medical images

The objective of this project is the development of model- and knowledge-based methods for shape analysis as well as automatic 3D-segmentation of diagnostically and therapeutically relevant objects in medical image volumes. The incorporation of a priori knowledge about the shape and the context of image structures allows a more robust segmentation of structures that feature weak edges or inhomogeneous intensities.

The probabilistic approach developed here is aimed to augment the possibilities of representing the natural 3D shape variability of anatomical structures in statistical shape models and to allow an optimal exploitation of shape information even in training data containing few observations. The idea is to integrate the probabilistic model into a flexible segmentation algorithm that should be able to deal with complex segmentation problems as non-spherical topologies or multi-object segmentation.

The new methods are evaluated on clinically relevant segmentation problems in the fields of radiation therapy and computer-aided intervention planning.

Probabilistic statistical shape model of the kidney. (a) shows the mean shape. (b-e) show the shape variations according to the first (b,c) and second (d,e) variation mode.

The project is funded by Deutsche Forschungsgemeinschaft (DFG: HA 2355/7-1).

Selected Publications:

  1. Hufnagel, H., Pennec, X., Ehrhardt, J., Handels, H. and Ayache, N. (2007). Shape Analysis Using a Point-Based Statistical Shape Model Built on Correspondence Probabilities. Proceedings of the MICCAI'07: 959-967.
  2. Hufnagel, H., Pennec, X., Ehrhardt, J., Ayache, N. and Handels, H. (2008). "Generation of a Statistical Shape Model with Probabilistic Point Correspondences and EM-ICP." International Journal for Computer Assisted Radiology and Surgery (IJCARS) 2(5): 265-273.
  3. Hufnagel, H., Ehrhardt, J., Pennec, X., Ayache, N. and Handels, H. (2009a). "Computing of Probabilistic Statistical Shape Models of Organs Optimizing a Global Criterion." Methods of Information in Medicine 48(4): 314-319.
  4. Hufnagel, H., Ehrhardt, J., Pennec, X., Schmidt-Richberg, A. and Handels, H. (2009b). Level Set Segmentation Using a Point-Based Statistical Shape Model Relying on Correspondence Probabilities. Proc. of MICCAI Workshop Probabilistic Models for Medical Image Analysis (PMMIA'09): 34-44.
  5. Hufnagel, H., Ehrhardt, J., Pennec, X. and Handels, H. (2009c). Application of a Probabilistic Statistical Shape Model to Automatic Segmentation. World Congress on Medical Physics and Biomedical Engineering, WC 2009, München: 2181-2184.
  6. Hufnagel, H., Ehrhardt, J., Pennec, X., Schmidt-Richberg, A. and Handels, H. (2010a). Coupled Level Set Segmentation Using a Point-Based Statistical Shape Model Relying on Correspondence Probabilities. Proc. SPIE Symposium on Medical Imaging 2010: 6914 6914T6911-6914T6918.

Project Team:

Dipl.- Inf. Heike Hufnagel
Dr. Jan Ehrhardt
Prof. Dr. Heinz Handels

Cooperation Partners:

Prof. Dr. Nicholas Ayache
Dr. Xavier Pennec
INRIA, Institut National de Recherche en Informatique et en Automatique, Epidaure Group, Sophia Antipolis Cedex, Frankreich


Created at September 13, 2010 - 11:37am. Last modified at June 30, 2014 - 11:43am.


Program of Study

Study Medical Informatics
at the University of Lübeck

read more ...


Susanne Petersen

Tel+49 451 3101 5601
Fax+49 451 3101 5604

Ratzeburger Allee 160
23538 Lübeck