Final Report Summary - BEATING HEART (Beating Heart) AbstractThe main objective of this project was to develop a laser scanning imaging platform capable of performing in-vivo real-time microscopy in the beating heart (and other organs as consequence) which physiological movement hampers intaviatal imaging at subellular resolution.The main goal is divided in specific objectives: 1) Developing the imaging system Objective: Studying mechanical stabilizers capable to stabilize the heart and other organs by reducing movement and introducing reproducibility of the measurements. 2) Developing and applying the computational methods for extracting data and modeling responses Objective: developing pacing protocols combined with triggering strategies and the required algorithms for imaging, extracting or processing information for microscopically imaging the heart intravitally.3) Developing real-time cardio-respiratory gating to enable multidimensional imaging of tissue microstructure and function Objective: studying contractile myocardial cells and developing of methods to image the beating heart in three dimensions and extending to four dimensions. 4) Building an analogue of the imaging system within the incoming institution. Objective: to build a similar imaging system with same imaging capabilities and to develop algorithms to perform specific investigations in beating heart imaging.- a description of the work performed since the beginning of the project, and a description of the main results achieved so far:Outgoing phase. Both active and passive stabilizers were implemented tested and used. Optimization process was adopted in order to achieve the best performance in terms of stabilization and in terms of compatibility to minimally perturb the heart functions. Testing was achieved by using phantoms made of fluorescent beads embedded in agar for simulating the biological tissue while a vibrating source was used for simulating the organ movement. The designed stabilizers provide optimal stabilization because of significant reduction of local motion (around the imaged area) and also because they introduce repeatability of the measurements over time. Different acquisition protocols have been implemented. Such protocols allowed us to compensate for the residual movement that the mechanical stabilizers cannot suppress. Also computational methods have been developed to overcome encountered issues: since the acquired signal may exceed the detector PMT range, a high dynamic range method for fluorescence imaging has been developed; for dealing with low fluorescence signal, denoise algorithms have been implemented instead. Moreover image processing algorithms and methods, to extract information from data, have been designed too. Cardiovascular imaging was enabled by the developed platform. In order to evaluate the effectiveness of the system concerning living applications, stabilization was tested with living animals. Retrospective and prospective imaging was effective in intravital cardiac imaging. Experiments with microbeads revealed that the stabilization method used in combination with gating permits to reach resolution below the one required to imaging at cellular level. Such a resolution enabled imaging of myocyte contractility. Besides study the pharmacokinetics of fluorescently labeled drugs in the beating heart were possible with promising preliminary results. Specific methods to visualize and quantify drugs’ pharmacokinetics based on Principal and Independent Component Analysis algorithms were partially developed (Fig.1). Ingoing phase. In order to overcome the lack of presence of any triggering signal in the commercial imaging microscope present within the ingoing institution a custom made electronic device was implemented. Scattered light from the microscope lasers was acquired by a photodetector and used to infer the linescan waveform timing. The system setup is shown in fig. 2.Testing was performed with phantoms in a similar manner as for the outgoing institution. Computational methods were tested and fully functionality verified. The setup was proved valid for intravital imaging of the heart. The imaging animal was anesthetized and a mechanical ventilator was used to provide artificial respiration after thoracotomy. Fluorescently labeled compounds were injected for capillary network and leukocytes staining. Heart was stabilized mechanically and the animal heart imaged. Computational methods are essential in biological imaging as images contain information that needs to be extracted and quantified before performing further analysis. This is a time consuming procedure if performed manually. Leukocytes transmigration study is an example of application that benefits from automatic segmentation algorithms. We then implemented a computational method based on a machine learning paradigm to segment leukocytes and useful for tracking and analysis. The machine learning algorithm consisted of two main parts: training and testing. Raw image and its features were used as input data. The classifier was then trained based on human classification and tested on different data. Moreover, in order to gain information about drug binding at cellular level with potential applications on the beating heart imaging we also implemented a denoise algorithm for overcoming the complex problem of noise in this kind of imaging (Intravital Fluorescence Anisotropy Imaging). -Expected final results and their potential impact and useEven if significant advancements have been made in the health care during the latest decades, cardiovascular diseases are remarkable cause of death among developed countries with high social costs. The costs may even increase significantly if after myocardial infarction event, the patient survives. The developed technology will help to deepen our understanding about the biological processes related to the cardiac physiology at the cellular level and its function. In fact the higher resolution provided by microscopy over other existing technologies such as MRI, CT even though they also aim to similar objectives, they are (at the current state of the art) limited by the required resolution. The potential for in-vivo real-time microscopy imaging of the beating heart may shed new light to in vivo cell physiology pathways and help gaining new insights for novel drug discoveries. Computational methods may also fasten analysis or increase the accuracy of findings as time consuming image processing may be automatized. The gaining of new knowledge may have positive consequences in terms of prevention, diagnosis, treatments which represent priority societal challenges. Contacts:prof. Andrea Sbarbati, MD, PhD, Department of Neurological, Biomedical and Movement Sciences, University of Verona, strada Le grazie 8 - 37134 Verona, e-mail: andrea.sbarbati@univr.itPaolo Fumene Feruglio, PhD, Department of Neurological, Biomedical and Movement Sciences, University of Verona, strada Le grazie 8 - 37134 Verona, e-mail: paolo.fumeneferuglio@univr.itprof. Claudio Vinegoni, PhD, Center for Systems Biology, MGH-Harvard University, 185 Cambridge StreetBoston, MA 02114, e-mail: cvinegoni@mgh.harvard.edu final1-logo.png Related documents final1-figures-final-report.docx
