I design and build advanced optical microscopy systems for biological research, such as qMAPP, quantitative microscope for amplitude, phase, and polarization. Applications include quantitative 3D reconstruction of transparent (unlabeled) cells and tissue for biomarker discovery, label-free quantitative imaging of live cells for research and high-throughput high-content screening.
I am developing computational optical imaging systems, with emphasis on biological imaging. Essential tasks include: (i) improved methods for extracting information about the imaged target (e.g. multi-modal imaging) and (ii) new algorithms capable of using that information for automated target detection, recognition and interpretation. The computational and quantitative aspects of optical imaging are inextricably connected: only a complete and quantitative representation of a sample/target allows extensive computational approaches to image formation and interpretation. In the computational imaging paradigm, the “image” is no longer a visual representation (e.g. Rayleigh's criterion for resolution). Instead, it becomes the mathematical representation of the target's physical features in the data provided by the imaging system. My research is motivated by the following observations:
(O1)Many biological samples, including live cells and tissues, are transparent, with most information about their morphology contained by their phase image, thus invisible to bright field microscopy but accessible trough quantitative phase imaging.
(O2)Transparent cell and tissue morphology is important in many areas of biological research including: live cell dynamics, evaluation of cellular mass distribution, cell motility and cell function.
(O3)Imaging, as a research tool, relies heavily on image processing, which, due to lack of sufficient appropriate data, is often performed sub-optimally, generating artifacts. No existing imaging method can individually fulfill the requirements imposed by diverse situations (see also (O1)). Combining data provided by multiple imaging methods, using appropriate mathematical processing, will enhance individual images and generate additional useful information.
(O4)Efficient and rapid progress in biological research using optical microscopy requires a significant increase in the extent of using advanced automated data processing and interpretation. With the exception of fluorescence and pure attenuation (amplitude) images, the data provided by current microscopes is not amenable to extensive computer processing because of its incorrect representation of the physical properties of a transparent target.
These observations indicate that the biggest obstacle to the development of computational optical imaging for biological research is the robust (easy, direct, routine) and reliable (accurate, quantitative) imaging of phase and polarization. Quantitative phase microscopy has been recently recognized as an important new imaging tool leading to a number of developments in the field. Despite these developments, phase measurements remain highly susceptible to environmental factors (e.g. vibrations) leaving the problem of robust phase acquisition unsolved and severely limiting the translation to clinical applications.
Over the course of my graduate studies and professional career, I have been working on a solution to this problem. I invented the world's first high resolution, robust and accurate full-field (phase, amplitude, and polarization) sensor, the sampling field sensor (SFS). Over the past 5 years I developed qMAPP, a quantitative microscope for amplitude, phase, and polarization. Unlike other quantitative phase microscopes, qMAPP is capable of robust phase and polarization acquisition (e.g. imaging unlabeled transparent cells and tissue) at high resolution and sensitivity. Another unique feature is the simultaneous acquisition of both phase and polarization data.
I am currently exploring applications like quantitative 3D reconstruction of transparent (unlabeled) cells and tissue for biomarker discovery and label-free quantitative imaging of live cells for research and high-throughput high-content screening.