Computer Aided Diagnosis
From UMCN Radiology
Contents |
[edit] About
Computer Aided Diagnosis (CAD) is directed towards the development and application of new methods for computerized analysis of images to help radiologists in performing and improving diagnostic tasks.
Expertise in the area of medical imaging is essential for a Radiology Department that aims for recognition in scientific research. To make optimal use of the large volumes of image data produced by modern scanners, advanced methods for analysis and visualization are required. Besides this there is a growing need for computer use in the detection of disease in screening programs. With pioneering work in the field of computer aided detection of breast cancer in screening mammograms this line of research at the department gained world-wide recognition. Currently, computer aided diagnosis is one of the most rapidly expanding fields in radiology, covering a broad field of applications. In the radiology department research is directed at improvement of detection, diagnostics and treatment in oncological applications.
In the coming years, research in the area of computer-assisted detection and diagnosis of breast cancer will be continued. Three modalities have a central role: Mammography, MRI and Ultrasound. With respect to mammography and ultrasound, the research will focus on improvement of detection in the Dutch population screening study. The existing cooperation with the Dutch Reference Center for Breast Cancer (LRCB) shall be intensified. Due to playing an active role in the digitization of the Dutch population screening study, unique possibilities for research are created. Central archiving in the Netherlands of the screening mammograms allows for new possibilities of data-mining and further development of detection methods. Digitization also allows the possibility for in-service monitoring and training during screening. This offers a good starting point for experimental validation of computer-assisted detection.
One of the department’s focus areas in research is the application of MRI for the detection of breast cancer in a high-risk population and for diagnosis of prostate cancer. This research can only be achieved with the availability of advanced sequences and quantitative image processing techniques. By making use of pharmacokinetic models, it is possible to determine characteristic tissue parameters and to increase diagnostic specificity. A clinical application has been developed wherein new methods can be prospectively studied in the coming years. Further developments shall be directed towards detection and objective measurement of the likelihood of malignancy of detected abnormalities, and of integration with MR spectroscopy. Research shall be concomitantly developed for an effective method for prostate cancer screening in a high-risk population.
In numerous clinical applications, images of various modalities are used and it is more frequently required to fuse these images. In a previous project an image fusion application was developed, which is now applied for diagnosis and for planning in radiotherapy. Depending on the need, this application will be expanded and further developed for sophisticated data registration.
In the coming years, new developments are anticipated in neuroradiology and interventional radiology. Initiatives in these areas will be actively supported.
[edit] People
- Karssemeijer N Head
- Huisman HJ Staff
- Velikova M PostDoc
- Snoeren PR PostDoc
- Kallenberg MGJ PhD
- Samulski MRM PhD
- Hupse A.M.G. PhD
- Hambrock T PhD
- Vos P.C. PhD
- van Schie, Guido PhD
- Miranda Zijp PhD
- Aliaksei Makarau PhD
- Oscar Debats PhD
[edit] Projects
- Computer Aided Detection of Masses in Mammograms
- Analysis of microcalcification clusters
- Soft Copy Reading of Mammograms
- Breast Ultrasound
- Full Field Digital Mammography (FFDM)
- Multimodality Image Registration
- Computer aided diagnosis of breast MR
- Computer aided diagnosis of prostate MR
- Bayesian Decision Support in Medical Screening
- Highly Accurate Breast Cancer Diagnosis through Integration of Biological Knowledge, Novel Imaging Modalities, and Modelling
- Computer aided detection of lymph nodes in MR lymphography
[edit] PhD theses
- W. Veldkamp Computer aided characterization of microcalcification clusters in mammograms
- G. te Brake Computer aided detection of masses in digital mammograms
- S. van Engeland Detection of mass lesions in mammograms by using multiple views
- S. Timp Analysis of temporal mammogram pairs to detect and characterise mass lesions
