Projects / Programmes
Image registration in image-guided medical interventions
| Code |
Science |
Field |
Subfield |
| 2.06.00 |
Engineering sciences and technologies |
Systems and cybernetics |
|
| Code |
Science |
Field |
| T111 |
Technological sciences |
Imaging, image processing |
medical images; image processing and analysis; image registration; image-guided interventions; surgery; radiotherapy; radiosurgery; chemotherapy; minimally invasive interventions
Researchers (8)
| no. |
Code |
Name and surname |
Research area |
Role |
Period |
No. of publicationsNo. of publications |
| 1. |
25608 |
PhD Marko Bukovec |
Computer science and informatics |
Researcher |
2008 - 2010 |
0 |
| 2. |
25528 |
PhD Miran Burmen |
Systems and cybernetics |
Researcher |
2009 - 2011 |
0 |
| 3. |
29559 |
PhD Jaka Katrašnik |
Systems and cybernetics |
Junior researcher |
2009 - 2011 |
0 |
| 4. |
15678 |
PhD Boštjan Likar |
Systems and cybernetics |
Researcher |
2009 - 2011 |
0 |
| 5. |
27519 |
PhD Primož Markelj |
Systems and cybernetics |
Junior researcher |
2009 - 2010 |
0 |
| 6. |
06857 |
PhD Franjo Pernuš |
Systems and cybernetics |
Head |
2008 - 2011 |
0 |
| 7. |
28465 |
PhD Žiga Špiclin |
Systems and cybernetics |
Junior researcher |
2009 - 2011 |
0 |
| 8. |
23404 |
PhD Tomaž Vrtovec |
Systems and cybernetics |
Researcher |
2008 - 2011 |
0 |
Organisations (1)
Abstract
Medical imaging is currently undergoing rapid development with a strong emphasis being placed on the use of imaging technology to render medical interventions (surgery, radiotherapy, radiosurgery, biopsy, etc.) less and less invasive and to improve the accuracy of conventional methods. Visualization during an intervention is limited, as the surgeon, radiotherapist or interventionalist cannot see beyond the exposed surfaces. Within the limits of the surgical opening, the exposed visible field lacks depth clues to comprehend the entire anatomy of interest, making it difficult to perform the procedure as accurately as desired. The approach that has the potential to decrease the invasiveness and increase the accuracy of a procedure is intra-operative image guidance or surgical navigation, which complements the physician’s view of the exposed surface by pre- or/and intra-operatively acquired three-dimensional (3D) images showing the scene below the surface. Under image guidance, the surgeon, radiotherapist, radio-surgeon, or chemotherapist can minimally invasively and accurately direct a variety of mechanical probes, energy sources, or controlled-release chemicals to targets avoiding vulnerable structures on the way. Regardless of the application, the conceptual design of image-guided interventional (IGI) systems is similar, although different applications require additions or specific modifications of individual components. The major components of an IGI system are: pre- and intra-operative images (imaging devices), image-to-patient registration, three-dimensional tracking, and simultaneous display of images, preoperative plan, and surgical tools. Preoperatively acquired 3D computed tomography (CT) or magnetic resonance (MR) images provide information important for the visualization, localization, and targeting of the three-dimensional (3D) anatomy of interest. The raw images and the anatomic models derived by image segmentation support preoperative planning to define and optimize access strategies and to simulate planned interventions. As the preoperative images, plan and simulation are part of the image-guided procedure that will be carried out in the operating room these data have to be integrated with the corresponding anatomy of the patient during intervention. The link between preoperative images and intra-operative physical space of the patient is established by image-to-patient registration (fusion, matching). To be suitable for a clinical application, a registration algorithm must satisfy several requirements. These requirements concern registration accuracy, robustness, computational time and complexity, and invasiveness of intra-operative data acquisition. Although a variety of methods have been proposed in the past, registration of pre- and intra-operative images of different modalities and dimensions is still a vital field of research in applied medical image processing and analysis. The problem of accurate and robust registration is far from being solved and therefore most existing registration algorithms still do not allow clinical application. Improvement of the performance of registration algorithms is therefore mandatory. In the proposed research project we will focus on the development of novel image registration methods that will have the potential to be used in image-guided interventions. Before an IGI system is put into clinical use, it must undergo rigorous validation, which includes validation of individual components, validation of the overall system and a study of how uncertainties propagate through the entire IGI process. Prerequisites for validation of an IGI component, like 3D-2D registration, is standardization of the validation methodology, which should include design of validation data sets, definition of corresponding “ground truth” and its accuracy, validation protocol, and design of validation metrics. Therefore, an important part of planned research efforts will be devoted to the development of
Significance for science
Based on the results of this basic research project which are so far summarized in 7 papers in peer reviewed journals, most of them in journals in the top 25% of journals in a research field, and numerous presentations at conferences, we expect that the results of this project will have an strong impact on the science in the field of automated medical image analysis. The group is already recognized as one of the leading groups in registration methods for image-guided medical interventions. Besides the impact on science, we believe that our research results have a potential to contribute to novel technologies for innovative navigation in image-guided procedure.
The development of efficient methods for the registration of pre- and intra-interventional images, which is the key enabling technology of image-guided interventions, requires extensive research at least on three domains: (a) development of accurate and efficient registration methods, (b) development of standardized validation protocols that is shared between researchers, and (c) extensive validation and comparison of several state-of-the-art methods. We believe that within this project we have scientifically contributed to each of the above mentioned domains.
(a) We have developed a promising registzration method that is based on intensity gradients:
• MARKELJ, Primož, TOMAŽEVIČ, Dejan, PERNUŠ, Franjo, LIKAR, Boštjan. Robust gradient-based 3-D/2-D registration of CT and MR to X-ray images. IEEE Transactions on Medical Imaging, Dec. 2008, vol. 27, no. 12, str. 1704-1714
• MARKELJ, Primož, TOMAŽEVIČ, Dejan, PERNUŠ, Franjo, LIKAR, Boštjan. Robust 3-D/2-D registration of CT and MR to X-ray images based on gradient reconstruction. International journal of computer assisted radiology and surgery, 2008, vol. 3, no. 6, str. 477-483
(b) We have developed two (one in collaboration with the Medical University Vienna) standardized validation protocols
• MARKELJ, Primož, LIKAR, Boštjan, PERNUŠ, Franjo. Standardized evaluation methodology for 3D/2D registration based on the Visible Human data set. Med. phys. (Lanc.), Sep. 2010, vol. 37, no. 9, str. 4643-4647
• PAWIRO, S. A., MARKELJ, Primož, PERNUŠ, Franjo. Validation for 2D/3D registration I : a new gold standard data set. Med. phys. (Lanc.), Mar. 2011, vol. 38, no. 3, str. 1481-1490
(c) Numerous registration methods were extensively validated and compared:
• GENDRIN, C., MARKELJ, Primož, LIKAR, Boštjan, PERNUŠ, Franjo. Validation for 2D/3D registration II : the comparison of intensity- and gradient-based merit functions using a new gold standard data set. Med. phys. (Lanc.), Mar. 2011, vol. 38, no. 3, str. 1491-1502
Besides, we have published a thorough review of registration methods:
• MARKELJ, Primož, TOMAŽEVIČ, Dejan, LIKAR, Boštjan, PERNUŠ, Franjo. A review of 3D/2D registration methods for image-guided interventions. Med. image anal. (Print), 2010, vol. , no. , str. 1-20
and helped our colleagues in Vienna to efficiently implement their registration algorithm:
• FIGL, M., BLOCH, C., GENDRIN, C., WEBER, C., PAWIRO, S. A., HUMMEL, J., MARKELJ, Primož, PERNUŠ, Franjo, BERGMANN, Helmar, BIRKFELLNER, W. Efficient implementation of the rank correlation merit function for 2D/3D registration. Phys. Med. Biol., Oct. 2010, vol. 55, no. 19, str. 465-471
Significance for the country
The ultimate success of a new medical technology, in this case part of the technology (image registration) for image-guided interventions, is judgeded by its impact on clinical practice and also through its contributions to scientific knowledge. The following specific benefits for Slovenia can easily be identified:
Benefits in terms of the results: The results of the research project had a direct impact on the state of science in Slovenia in the fields of biomedical engineering and image-guided interventions. Above all they are giving the field of biomedical engineering (BME) a momentum, as BME is becoming an important research and development field and in the near future also a field of higher education. The large number of publications, in respect to the funding of the project, in top peer reviewed journals and international conferences helped to increase the reputation of Slovenia as the country with high-quality research. The number and quality of publications are directly and immediately reflected in: editorial boards memberships, membership in PC committees of international conferences, invited talks at several universities and companies abroad, memberships on PhD evaluation and defence committees at recognized universities, interest of foreign students in PhD studies at our university (during the project we have attracted three such students), invitations to participate at common projects, visits of recognized experts in the field, interest of foreign companies in the field of medical imaging for our work, etc.
Human resources development: The abilities and skills of the researchers fully or partially connected to the research project have significantly improved. Besides the direct transfer of knowledge from academia to industry, knowledge is also transferred through the well trained and highly motivated young researchers who worked on the project and decided to find a job in industry after obtaining their degree. Because of the highly relevant and interesting research topic and the expertise of the research team, it is expected that future projects will attract new talented researchers.
Strengthening of the institutional capacities: The research project contributed to the capability of acquiring new and reconditioning, improving and renovating the existing elements of hardware, software, literature, and administrative and management systems.
Increased and more effective collaboration: Working relationships between different individuals and institutions, whether they were directly related to the research or not, established new and strengthened old collaborations (for instance with Medical University Vienna), which might lead to more effective future collaborations.
Most important scientific results
Annual report
2008,
2009,
final report,
complete report on dLib.si
Most important socioeconomically and culturally relevant results
Annual report
2008,
2009,
final report,
complete report on dLib.si
