Wouter Caarls

Tumor-Targeted Quantum Dots Can Help Surgeons Find Tumor Boundaries

Despite surgical advances and recent progress in adjuvant therapies, the prognosis for patients with malignant brain tumors such as glioblastoma multiforme has remained poor, and the neurological deterioration suffered by most patients as a consequence of tumor progression is dramatic and severe. In addition, malignant brain tumors have > 95% recurrence close to the primary site of initial resection. Unfortunately, standard imaging techniques do not permit the intraoperative identification of individual or small clusters of residual tumor cells, precluding their selective removal while sparing the surrounding normal brain tissue. In this report, we show that quantum dots (QDs) coupled to epidermal growth factor (EGF) or anti-EGF receptor (EGFR, Her1) specifically and sensitively label glial tumor cells in cell culture, glioma mouse models, and human brain-tumor biopsies. A clear demarcation between brain and tumor tissue at the macroscopic as well as the cellular level is provided by the fluorescence emission of the QDs.

Specific Visualization of Glioma Cells in Living Low-Grade Tumor Tissue

Background The current therapy of malignant gliomas is based on surgical resection, radio-chemotherapy and chemotherapy. Recent retrospective case-series have highlighted the significance of the extent of resection as a prognostic factor predicting the course of the disease. Complete resection in low-grade gliomas that show no MRI-enhanced images are especially difficult. The aim in this study was to develop a robust, specific, new fluorescent probe for glioma cells that is easy to apply to live tumor biopsies and could identify tumor cells from normal brain cells at all levels of magnification. Methodology/Principal Findings In this investigation we employed brightly fluorescent, photostable quantum dots (QDs) to specifically target epidermal growth factor receptor (EGFR) that is upregulated in many gliomas. Living glioma and normal cells or tissue biopsies were incubated with QDs coupled to EGF and/or monoclonal antibodies against EGFR for 30 minutes, washed and imaged. The data include results from cell-culture, animal model and ex vivo human tumor biopsies of both low-grade and high-grade gliomas and show high probe specificity. Tumor cells could be visualized from the macroscopic to single cell level with contrast ratios as high as 1000: 1 compared to normal brain tissue. Conclusions/Significance The ability of the targeted probes to clearly distinguish tumor cells in low-grade tumor biopsies, where no enhanced MRI image was obtained, demonstrates the great potential of the method. We propose that future application of specifically targeted fluorescent particles during surgery could allow intraoperative guidance for the removal of residual tumor cells from the resection cavity and thus increase patient survival.

Characterization of Coupled Ground State and Excited State Equilibria by Fluorescence Spectral Deconvolution

Fluorescence probes with multiparametric response based on the relative variation in the intensities of several emission bands are of great general utility. An accurate interpretation of the system requires the determination of the number, positions and intensities of the spectral components. We have developed a new algorithm for spectral deconvolution that is applicable to fluorescence probes exhibiting a two-state ground-state equilibrium and a two-state excited-state reaction. Three distinct fluorescence emission bands are resolved, with a distribution of intensities that is excitation-wavelength-dependent. The deconvolution of the spectrum into individual components is based on their representation as asymmetric Siano-Metzler log-normal functions. The application of the algorithm to the solvation response of a 3-hydroxychromone (3HC) derivative that exhibits an H-bonding-dependent excited-state intramolecular proton transfer (ESIPT) reaction allowed the separation of the spectral signatures characteristic of polarity and hydrogen bonding. This example demonstrates the ability of the method to characterize two potentially uncorrelated parameters characterizing dye environment and interactions.

Biological applications of an LCoS-BASED PROGRAMMABLE ARRAY MICROSCOPE (PAM)

We report on a new generation, commercial prototype of a programmable array optical sectioning fluorescence microscope (PAM) for rapid, light efficient 3D imaging of living specimens. The stand-alone module, including light source(s) and detector(s), features an innovative optical design and a ferroelectric liquid-crystal-on-silicon (LCoS) spatial light modulator (SLM) instead of the DMD used in the original PAM design. The LCoS PAM (developed in collaboration with Cairn Research, Ltd.) can be attached to a port of a(ny) unmodified fluorescence microscope. The prototype system currently operated at the Max Planck Institute incorporates a 6-position high-intensity LED illuminator, modulated laser and lamp light sources, and an Andor iXon emCCD camera. The module is mounted on an Olympus IX71 inverted microscope with 60-150X objectives with a Prior Scientific x,y, and z high resolution scanning stages. Further enhancements recently include: (i) point- and line-wise spectral resolution and (ii) lifetime imaging (FLIM) in the frequency domain. Multiphoton operation and other nonlinear techniques should be feasible. The capabilities of the PAM are illustrated by several examples demonstrating single molecule as well as lifetime imaging in live cells, and the unique capability to perform photoconversion with arbitrary patterns and high spatial resolution. Using quantum dot coupled ligands we show real-time binding and subsequent trafficking of individual ligand-growth factor receptor complexes on and in live cells with a temporal resolution and sensitivity exceeding those of conventional CLSM systems. The combined use of a blue laser and parallel LED or visible laser sources permits photoactivation and rapid kinetic analysis of cellular processes probed by photoswitchable visible fluorescent proteins such as DRONPA.

Comparison of extremum seeking control algorithms for robotic applications

The purpose of this paper is to help engineers and researches to choose among the extremum seeking control (ESC) techniques for robotic applications such as object grasping, active object recognition and viewpoint optimization. These techniques are categorized into five main groups: Sliding mode ESC, neural network ESC, approximation based ESC, perturbation based ESC and adaptive ESC. These groups are explained briefly by stressing their working principles and the effect of the parameters. Then, the techniques are compared with respect to their robustness to noise and system dynamics by simulations. In conclusion, we propose the usage of the approximation based methods when the noise level is negligible. When noise is present, the neural network based optimizers are a better choice thanks to their hysteresis functions. However, if the system has both high noise and dynamic effects, then the perturbation based method is preferable since large motions provide robustness to noise and smooth references generated by the algorithm are less likely to cause instability. An application example is also given on texture density maximization.

Minimizing light exposure with the programmable array microscope.

The exposure of fluorophores to intense illumination in a microscope often results in photobleaching and phototoxicity, thus constituting a major limiting factor in time lapse live cell or single molecule imaging. Laser scanning confocal microscopes are particularly prone to this problem, inasmuch as they require high irradiances to compensate for the inherently low duty cycle of point scanning systems. In the attempt to maintain adequate speed and signal‐to‐noise ratios, the fluorophores are often driven into saturation, thereby generating a nonlinear response.

Fluorescence recovery after photobleaching and photoconversion in multiple arbitrary regions of interest using a programmable array microscope.

Photomanipulation (photobleaching, photoactivation, or photoconversion) is an essential tool in fluorescence microscopy. Fluorescence recovery after photobleaching (FRAP) is commonly used for the determination of lateral diffusion constants of membrane proteins, and can be conveniently implemented in confocal laser scanning microscopy (CLSM). Such determinations provide important information on molecular dynamics in live cells. However, the CLSM platform is inherently limited for FRAP because of its inflexible raster (spot) scanning format. We have implemented FRAP and photoactivation protocols using structured illumination and detection in a programmable array microscope (PAM). The patterns are arbitrary in number and shape, dynamic and adjustable to and by the sample characteristics. We have used multispot PAM–FRAP to measure the lateral diffusion of the erbB3 (HER3) receptor tyrosine kinase labeled by fusion with mCitrine on untreated cells and after treatment with reagents that perturb the cytoskeleton or plasma membrane or activate coexpressed erbB1 (HER1, the EGF receptor EGFR). We also show the versatility of the PAM for photoactivation in arbitrary regions of interest, in cells expressing erbB3 fused with the photoconvertible fluorescent protein dronpa. dronpa. Microsc. Res. Tech., 2009.

Algorithmic skeletons for stream programming in embedded heterogeneous parallel image processing applications

Algorithmic skeletons can be used to write architecture independent programs, shielding application developers from the details of a parallel implementation. In this paper, we present a C-like skeleton implementation language, PEPCI, that uses term rewriting and partial evaluation to specify skeletons for parallel C dialects. By using skeletons to control the iteration of kernel functions, we provide a stream programming language that is better tailored to the user as well as the underlying architecture. Skeleton merging allows us to reduce the overheads usually associated with breaking an application into small kernels. We have implemented an example image processing application on a heterogeneous embedded prototype platform consisting of an SIMD and ILP processor, and show that a significant speedup can be achieved without requiring knowledge of data parallel processing.

Skeletons and Asynchronous RPC for Embedded Data and Task Parallel Image Processing*This work is supported by the Dutch government in their PROGRESS research program under project EES.5411.

Developing embedded parallel image processing applications is usually a very hardware-dependent process, often using the single instruction multiple data (SIMD) paradigm, and requiring deep knowledge of the processors used. Furthermore, the application is tailored to a specific hardware platform, and if the chosen hardware does not meet the requirements, it must be rewritten for a new platform. We have proposed the use of design space exploration [9] to find the most suitable hardware platform for a certain application. This requires a hardware-independent program, and we use algorithmic skeletons [5] to achieve this, while exploiting the data parallelism inherent to low-level image processing. However, since different operations run best on different kinds of processors, we need to exploit task parallelism as well. This paper describes how we exploit task parallelism using an asynchronous remote procedure call (RPC) system, optimized for low-memory and sparsely connected systems such as smart cameras. It uses a futures [16]-like model to present a normal imperative C-interface to the user in which the skeleton calls are implicitly parallelized and pipelined. Simulation provides the task dependency graph and performance numbers for the mapping, which can be done at run time to facilitate data dependent branching. The result is an easy to program, platform independent framework which shields the user from the parallel implementation and mapping of his application, while efficiently utilizing on-chip memory and interconnect bandwidth.

Distance metric approximation for state-space RRTs using supervised learning

The dynamic feasibility of solutions to motion planning problems using Rapidly Exploring Random Trees depends strongly on the choice of the distance metric used while planning. The ideal distance metric is the optimal cost of traversal between two states in the state space. However, it is computationally intensive to find the optimal cost while planning. We propose a novel approach to overcome this barrier by using a supervised learning algorithm that learns a nonlinear function which is an estimate of the optimal cost, via offline training. We use the Iterative Linear Quadratic Regulator approach for estimating an approximation to the optimal cost and learn this cost using Locally Weighted Projection Regression. We show that the learnt function approximates the original cost with a reasonable tolerance and more importantly, gives a tremendous speed up of a factor of 1000 over the actual computation time. We also use the learnt metric for solving the pendulum swing up planning problem and show that our metric performs better than the popularly used Linear Quadratic Regulator based metric.