Laboratory for Imaging Technology, Methods, and Systems (ITeMS)

Director: Dr. Nikolaos Deliolanis

The development of innovative in vivo imaging technology is a core requirement for progress in basic as well as applied biological research. The advancement from non-invasively generating 3D images of the morphologic structures using X-ray CT or MRI, over extracting functional information about vascularisation, metabolic activity and oxygenation using fMRI, PET, and SPECT, to the current development of methods quantifying and spatially resolving cellular and subcellular functions in vivo using novel contrast agents has spawned an abundance of important potential applications in preclinical as well as clinical research. The development of suitable imaging methods has always played a central part in the success of experimental in vivo studies, but with the challenging need to display entities like cell receptors, enzymatic activity, and gene expression in correlation to overall morphology and function, the important role of technological developments has ever increased.

The Laboratory for Imaging Technology, Methods, and Systems (ITeMS) is aiming at developing and enhancing modern preclinical imaging systems to broaden the application spectrum for biomedical research. We have a strong background in developing the experimental system to perform modern optical technologies such as fluorescence-mediated molecular tomography (FMT) and multispectral and time-resolved methods. Furthermore, we develop the theoretical models and approximations necessary in the development of enhanced reconstruction algorithms, also including potential a priori information from other sources. Last not least we are experienced to put established (pre-)clinical modalities such as XCT and MRI into service for enhanced data acquisition, image reconstruction and display. Some of the key areas of our research are outlined below:

Fluorescence Imaging and Tomography

Fluorescence-mediated Molecular Tomography (FMT) resolves the concentration and localization of fluorescent markers in-vivo. It is a fast evolving imaging modality, which provides both functional and molecular information on gene expression, protein function, elucidation of cellular pathways, and small-molecule interactions in living specimen. The use of FMT has been reported for in vivo imaging of cancer progression and therapy efficacy, cancerous tissue, myocardial infarction, as well as lung inflammation in murine models. Techincally, FMT works by exciting fluorescent sources using differently positioned external sources, and recorded using some kind of light detection system such as CCD cameras. The quality of reconstructed images depends heavily on the number of available source-detector combinations and the use of appropriate mathematical models.

Optical instrumentation and hardware development for FMT is an essential part of our effort. Members of our lab have actively participated in the development of several generations of fluorescence molecular tomography (FMT) systems, enhancing the performance of probe detection and simplifying experimental procedures for improved image quality and user acceptance (see Fig. 1). We strive to further drive this progress by adding multiple wavelengths, which can enable visualization of multiple functional and molecular processes simultaneously, time- and spectral resolution for improved differentiation of fluorescent sources, removal of autofluorescence and improved resolution, or enhanced tissue models for reconstruction.
                          

Fig 1: Three generations of FMT systems with according results. (A) A first generation system, employing a fixed pattern of light-guiding fibers for detection and excitation from 360° of view around a cylindrical geometry. The co-registration of a Gadolinium-enhanced MR image of a mouse and the corresponding FMT data for a Cathepsin-B probe in 9L gliosarcomas is shown. (B) A second generation system with planar geometry, using a CCD camera for detection. Experimental procedures required the mouse to be embedded in an optically matching fluid. The FMT image shows LBS induced lung inflammation imaged with an activatable MMP-probe. Spatial resolution is in the order of 1mm or better. (C) Third generation systems do not require matching fluids or fibers, yielding high image performance together with a rendering of the animal’s geometry. Shown are reconstructions of two implanted inclusions in a mouse. For these systems refer to Ntziachristos et al, Nature Medicine, Vol. 8, pp. 757-760 (2002), Ntziachristos et al., Nature Biotechnology, Vol. 23, pp. 313-20 (2005), Deliolanis et al, Optics Letters, Vol. 32, pp. 383-5 (2007).

Multimodality Imaging

Molecular and functional imaging modalities such as FMT on their own provide only limited insight into the complex living system under investigation, as they usually fail to retrieve the global context of the molecular or functional events observed: smaller morphological structures like vascularization, epithelial structures, the lymphatic system, as well as the organism as a whole. Future modalities will have to incorporate anatomic, functional, and molecular information to gain a holistic image of the investigated biological organisms and diseases.

ITeMS strives to incorporate the key modalities of medical imaging – CT, MRI, PET, SPECT – and molecular imaging modalities such as FMT into hybrid systems in a way that optimizes the information content of resulting images as well as resolution and sensitivity concerning the cellular processes observed.         

Fig 2: An murine adenocarcinoma model was imaged using a Cathepsine-B sensitive probe. Shown are three FMT slices of a mouse at different depths, a corresponding X-ray CT slice with two lung tumors visible in the image, and a fused three-dimensional rendering of the X-ray CT and FMT data sets.

Theory Development

Appropriate and accurate theoretical descriptions, as well as computationally efficient algorithms are a core requirement for all image reconstruction problems. This is especially true for optical techniques, where the chosen model of photon propagation has an enormous influence on image quality as well as computation time. The integration of several imaging modalities into one extends further the requirements, as the combination of hybrid data should preferentially not consist in an overlay of resulting images, but in a true combination of the acquired data to yield optimal imaging performance and diagnostic output.

We research time-efficient forward and inversion schemes for fluorescence molecular tomography (FMT) and diffuse optical tomography (DOT) and spectroscopy using different approximations to the radiation transport equation, such as the diffusion or Fokker-Planck equations. Furthermore we optimize different forward models for the different imaging systems developed. An important other topic we consider is the development of appropriate methods for optimization of the information content of the various measurements performed. A key component of our research are computationally efficient finite element models on arbitrary meshes for solving the differential equations involved.