Three dimensional mapping of turbid media by hyper spectral imaging

• Applicant: Professor Dr.-Ing. Michael Schmidt
• Subject Area: Production Automation and Assembly Technology
• Term: from 2017 to 2021
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 337270237

Non-invasive modalities in medical imaging diagnostics are of vital importance in order to minimize the risks for patients, particularly since many diseases necessitate repetitive imaging. Furthermore, preventing high costs from impeding regular examinations, fast and cost-efficient imaging modalities are needed. Ideally, these methods should also enable label free imaging, as labels bear potential side effects for the patients such as allergies. Furthermore, the development of suitable markers is a costly and time-consuming process. Hyperspectral imaging systems have the potential to enable fast, low-cost, non-invasive and label free imaging. Thus, they provide the opportunity to overcome all the mentioned challenges. Therefore, the goal of the planned research project is to compile and to investigate methods for the transformation of hyperspectral images into three-dimensional tissue images. Two mechanisms leading to a wavelength dependency of the penetration depth in tissue will be used for this: the wavelength dependent absorption in tissue constituents such as hemoglobin and water as well the wavelength dependent scattering behavior of the tissue. Within the scope of this project different transformation approaches are investigated and evaluated. As outcome, an optical non-invasive modality for the examination of malign near-surface tissue alterations is expected which also provides the doctor with a large field of view.Compared to established optical 3D reconstruction methods such as Optical Coherence Tomography, the new approach has the advantage of not only measuring and analyzing changes in refractive index and scattering, but also changes in absorptivity. Therefore, absorber concentration changes, such as the amount of blood in cancerous and healthy tissue or scattering concentration changes, this means the amount of scatterers in blood vessel and surrounding tissue, should be detectable. This combination of the detectability of scattering and absorber alterations and the reconstruction of the in-depth resolved images promises improved detection of malign subsurface alterations. The detection is simplified by the fact that the images of a specific depth are not convoluted with all other depths anymore. This should increase the contrast significantly and it should enable a more reliable diagnosis since possible disguises of tissue alterations are removed. Furthermore, since the imaging approach is contact free and uses standard white light sources for illumination, it will enable low-risk and inexpensive repetitive examinations of patients.

• DFG Programme: Research Grants