Cookie Hinweis

Wir verwenden Cookies, um Ihnen ein optimales Webseiten-Erlebnis zu bieten. Dazu zählen Cookies, die für den Betrieb der Seite notwendig sind, sowie solche, die lediglich zu anonymen Statistikzwecken, für Komforteinstellungen oder zur Anzeige personalisierter Inhalte genutzt werden. Sie können selbst entscheiden, welche Kategorien Sie zulassen möchten. Bitte beachten Sie, dass auf Basis Ihrer Einstellungen womöglich nicht mehr alle Funktionalitäten der Seite zur Verfügung stehen. Weitere Informationen finden Sie in unseren Datenschutzhinweisen .


Diese Cookies sind für die Funktionalität unserer Website erforderlich und können nicht deaktiviert werden.

Name Webedition CMS
Zweck Dieses Cookie wird vom CMS (Content Management System) Webedition für die unverwechselbare Identifizierung eines Anwenders gesetzt. Es bietet dem Anwender bessere Bedienerführung, z.B. Speicherung von Sucheinstellungen oder Formulardaten. Typischerweise wird dieses Cookie beim Schließen des Browsers gelöscht.
Name econda
Zweck Session-Cookie für die Webanalyse Software econda. Diese läuft im Modus „Anonymisiertes Messen“.

Diese Cookies helfen uns zu verstehen, wie Besucher mit unserer Webseite interagieren, indem Informationen anonym gesammelt und analysiert werden. Je nach Tool werden ein oder mehrere Cookies des Anbieters gesetzt.

Name econda
Zweck Measure with Visitor Cookie emos_jcvid
Externe Medien

Inhalte von externen Medienplattformen werden standardmäßig blockiert. Wenn Cookies von externen Medien akzeptiert werden, bedarf der Zugriff auf diese Inhalte keiner manuellen Zustimmung mehr.

Name YouTube
Zweck Zeige YouTube Inhalte
Name Twitter
Zweck Twitter Feeds aktivieren

Holistic data processing for surgical decision support

< Back to Research overview


Primary Liver Cancer is one of the most common cancer types worldwide. Each year, an estimated number of half a million patients develop cancer of the liver. At the same time, attributed to its blood filtering function, the liver is the second most common site for metastases, with some authors projecting a 40-50% chance of developing liver metastases from primary extrahepatical tumors. The treatment of liver tumors is thus of enormous clinical importance. However, the choice of treatment is usually not obvious, as it depends on a wide range of factors.

These include:

Patient-individual data, which includes information that can be extracted from medical images (e.g. tumor number, size and location), laboratory reports, genetic data and other sources of information (e.g. clinical reports, hospital databases etc). It can be related to the disease (e.g. diagnosis, recurrence, genetic profile of tumor etc.), the anatomy and physiology (e.g. liver cirrhosis) or general information (e.g. age, habits etc).

Factual knowledge which is written down in quotable sources (e.g. clinical guidelines, studies, educational books etc). As it is impossible for a single surgeon to be aware of all relevant information, (hospital-specific) clinical guidelines make the amount of studies manageable by deducing general recommendations about recurring cases. However, a highly condensed guideline cannot appropriately represent uncommon cases and complicated treatment plans, which is especially the case for multimodal treatments. As a result, guidelines typically give merely rough directions, taking into account only a fraction of patient-individual parameters (e.g. size and number of tumors), while detailed treatment decisions remain to the physicians in charge (e.g. whether a resection is performed in a one or two stage approach, if it is done open or laparoscopically, etc). Furthermore, once created, guidelines are static until updated again to reflect the current state of knowledge and usually do not incorporate the most recent findings.

Practical knowledge which results from experience. It comprises case knowledge that encompasses the ability to interpret patient individual data, form a prognosis and deduct implications for the treatment as well as expert knowledge about treatment options and their respective strengths, weaknesses and their practical application (e.g. one-stage or two-stage resection; radiotherapy performed as a palliative, curative, adjuvant or neoadjuvant treatment, etc). Due to the limitations of clinical guidelines in providing patient-individual optimal treatment plans (cf. previous paragraph), hospitals increasingly employ tumor boards consisting of a multidisciplinary team of experts (surgical oncology, medical oncology, radiation oncology etc.) that derive (possibly multi-modal) treatment plans, taking into account not only factual knowledge and patient data but in particular the practical knowledge contributed by the board members.

Conquering the complexity of this problem and making it accessible to the treating physician is still an unsolved research question and the primary scope of this project.


As illustrated in Fig. 1, our vision is based on holistically processing all the relevant data involved in the treatment planning process to support the physician with the relevant information at the right time and thus to facilitate, optimize and objectify clinical decision making. The core component is the knowledge base, which dynamically acquires the patient data and all formalized knowledge.

Figure 1: Vision of holistic data processing for computer-assisted tumor therapy

The treatment process comprises four stages.

  1. Data acquisition and diagnosis: All the relevant patient individual data is collected and converted into a computer interpretable format. Where possible, the system derives the data automatically from text or image sources.
  2. Multimodal treatment planning: Based on the acquired data, the system processes the patient-individual data and the information in the knowledge base to derive a (possible multi-modal) treatment plan.
  3. Navigated treatment: The system assists the surgeon by providing adequate means of intra-operative navigation where necessary (which may be modified dynamically due to intra-operative findings).
  4. Follow-up: In this phase, treatment outcome is documented, and the current case is fed back into the knowledge base.

In the context of the Heidelberg/Karlsruhe Transregional Collaborative Research Center 125, we recently introduced a new concept for modeling, storing and accessing all the data that is used throughout the process of patient diagnosis, treatment planning, therapy and follow-up [1]. As the long-term success of the project proposed will depend crucially on reliable annotations of large amounts of data, machine learning algorithms will be among the key enabling techniques for a number of subtasks, such as the extraction of relevant patient parameters or the classification of medical images. One of the major bottlenecks related to applying such algorithms, however, is the lack of training data, which can be attributed to the limited resources of medical experts. In recent work, we proposed an entirely new concept for medical image annotation based on crowdsourcing. In two pilot studies [2,7,8] we showed that anonymous online workers can generate image annotations that can compete with that of experts – but the crowd is orders of magnitude faster (and cheaper).
With respect to the navigated treatment (Step 3 in Fig. 1), we have further introduced a new concept for navigated insertion of markers for radiotherapy [3]. Following the Echotrack concept. These markers are positioned around the tumor prior to radiation therapy for accurate tracking of tumor movement.

Key collaborators

  • Prof. Markus Büchler, Dr. Beat Müller-Stich and Dr. Arianeb Mehrabi
    University of Heidelberg, Division of Visceral Surgery of the Department of General Surgery

  • Dr. Stefanie Speidel and Prof. Rüdiger Dillmann
    Karlsruhe Institute for Technology, Humanoids and Intelligence Systems Labs

  • Prof. Markus Büchler, Dr. Beat Müller-Stich and Dr. Hannes Kenngott
    University of Heidelberg, Division of Minimally-invasive Surgery of the Department of General Surgery

  • Dr. Christian Stock 
    University of Heidelberg, Department of Medical Biometry

  • Daniel Kondermann
    University of Heidelberg, Heidelberg Collaboratory for Image Processing



This work was supported by the German Research Foundation (DFG) as part of project A02 of the SFB/TRR 125 Cognition-Guided Surgery.

Selected publications

  1. Maier-Hein L, Vedula SS, Speidel S, Navab N, Kikinis R, Park A, Eisenmann M, Feussner H, Forestier G, Giannarou S, Hashizume M, Katic D, Kenngott H, Kranzfelder M, Malpani A, März K, Neumuth T, Padoy N, Pugh C, Schoch N, Stoyanov D, Taylor R, Wagner M, Hager GD, Jannin P. Surgical data science for next-generation interventions. Nature Biomedical Engineering. 2017;1(9):691.

  2. Katić D, Maleshkova M, Engelhardt S, Wolf I, März K, Maier-Hein L, Nolden M, Wagner M, Kenngott H, Müller-Stich BP, Dillmann R, Speidel S. What does it all mean? Capturing Semantics of Surgical Data and Algorithms with Ontologies. arXiv:170507747 [cs] [Internet]. 2017; Available from:

  3. März K, Hafezi M, Weller T, Saffari A, Nolden M, Fard N, Majlesara A, Zelzer S, Maleshkova M, Volovyk M, others. Toward knowledge-based liver surgery: holistic information processing for surgical decision support. International Journal of Computer Assisted Radiology and Surgery. 2015;10(6):749-59.

Selected awards

  • Philips/IPCAI Best Presentation award (2015)
    Keno März et al., for the paper, "Towards Knowledge-Based Liver Surgery - Holistic Information Processing for Surgical Decision Support".

  • Fachbereichspreis for the best Master Thesis (2014)
    Adrian Winterstein, for Masters Thesis "Ein ultraschallbasiertes Computerassistenzsystem mit integriertem elektromagnetischem Feldgenerator für die Leberchirurgie: Konzeption und Realisierung im Rahmen von MITK", awarded by the University of Darmstadt

  • BVM Best Scientific Work (2013)
    Keno März et al., "Navigierte ultraschallgeführte Leberpunktion mit integriertem EM Feldgenerator"

< Back to Research overview

to top
powered by webEdition CMS