Cancer
|

A quality-managed software platform for quantitative microstructural tissue characterization using diffusivity imaging for early detection of cancer (Acronym: ARGOS-Imaging)

Institution: Universitätsklinikum Erlangen
Applicant: PD Dr. med. Sebastian Bickelhaupt
Co-applicant: PD Dr. med. Klaus Maier-Hein (Deutsches Krebsforschungszentrum Heidelberg)
Funding line:
Translational Research
Image: Graphic

Breast cancer is the most common malignant tumor disease for women. Approximately one in eight females in Germany develops breast cancer during her lifetime. Diagnostic imaging procedures are essential in supporting the early detection and characterization of tumor-suspicious changes in the female breast.
This joint translational EKFS project of the Universitätsklinikum Erlangen (UKER) and the German Cancer Research Center Heidelberg (DKFZ) aims to translate a new modular diagnostic approach using innovative AI-assisted imaging techniques into a clinical evaluation by developing a comprehensive software-prototype. In contrast to established imaging techniques, the new approach does not require any contrast agent, X-rays or chest compression at all.

To achieve this, the so-called diffusion-weighted magnetic resonance imaging (MRI) technique, which scans tissue structure by mapping the molecular motion of water molecules, is adapted to the requirements of early oncological tissue diagnostics and combined with novel quality assurance and calibration procedures. The data acquired with this approach can subsequently be analyzed with multiparametric quantitative classification models based on correlates of organ-specific tissue microstructure and, in particular, also in combination with the latest machine learning and artificial intelligence methods to enable the detection and largely autonomous, radiologist-independent characterization of suspicious findings. In addition to the detection and characterization of suspicious finding, the technology also opens up potential fields of application for stratifying individual treatment pathways in existing oncological diseases and for the assessment of the individual risk of developing breast cancer in future.

To be able to evaluate the prospects and limitations of this innovative approach beyond the existing initial scientific study projects and to enable the investigations and clinical studies necessary for an in-depth evaluation of the diagnostic performance, the individual components (which are currently still being manually assembled in the scientific research environment) will be merged into a unique application prototype that can be used outside a pure research environment. Such an application prototype, which can also be operated in a clinical environment, will allow the method to become integratively implemented in a multi-center clinical setting and be comprehensively evaluated in terms of prospective scientific trials.

Image: Team