Minimizing light exposure with the programmable array microscope

An adaptive microscope for the imaging of biological surfaces

Scanning fluorescence microscopes are now able to image large biological samples at high spatial and temporal resolution. This comes at the expense of an increased light dose which is detrimental to fluorophore stability and cell physiology. To highly reduce the light dose, we designed an adaptive scanning fluorescence microscope with a scanning scheme optimized for the unsupervised imaging of cell sheets, which underly the shape of many embryos and organs. The surface of the tissue is first delineated from the acquisition of a very small subset (~0.1%) of sample space, using a robust estimation strategy. Two alternative scanning strategies are then proposed to image the tissue with an improved photon budget, without loss in resolution. The first strategy consists in scanning only a thin shell around the estimated surface of interest, allowing high reduction of light dose when the tissue is curved. The second strategy applies when structures of interest lie at the cell periphery (e.g. adherens junctions). An iterative approach is then used to propagate scanning along cell contours. We demonstrate the benefit of our approach imaging live epithelia from Drosophila melanogaster. On the examples shown, both approaches yield more than a 20-fold reduction in light dose -and up to more than 80-fold- compared to a full scan of the volume. These smart-scanning strategies can be easily implemented on most scanning fluorescent imaging modality. The dramatic reduction in light exposure of the sample should allow prolonged imaging of the live processes under investigation.

Fluorescence microscopy datasets for training deep neural networks

BACKGROUND

Fluorescence microscopy is an important technique in many areas of biological research. Two factors that limit the usefulness and performance of fluorescence microscopy are photobleaching of fluorescent probes during imaging and, when imaging live cells, phototoxicity caused by light exposure. Recently developed methods in machine learning are able to greatly improve the signal-to-noise ratio of acquired images. This allows researchers to record images with much shorter exposure times, which in turn minimizes photobleaching and phototoxicity by reducing the dose of light reaching the sample.

FINDINGS

To use deep learning methods, a large amount of data is needed to train the underlying convolutional neural network. One way to do this involves use of pairs of fluorescence microscopy images acquired with long and short exposure times. We provide high-quality datasets that can be used to train and evaluate deep learning methods under development.

CONCLUSION

The availability of high-quality data is vital for training convolutional neural networks that are used in current machine learning approaches.

Semi-autonomous real-time programmable fluorescence lifetime segmentation with a digital micromirror device

Time-correlated single-photon counting (TCSPC) is the gold standard for performing lifetime spectroscopy in biological assays. Traditional fluorescence lifetime imaging (FLIM) using laser scanning microscopes are inherently slow due to point scanning all pixels in the field-of-view. Wide-field implementations of TCSPC spectroscopy using microchannel plates benefit from particularly fast acquisition times at the expense of temporal resolution, and are fundamentally limited by photon counting rates. Here, we introduce programmable lifetime imaging (PLI), combining the advantages of wide-field imaging using total internal reflection excitation with state-of-the-art TCSPC detector technology for accurate lifetime determination in an object-oriented manner using a digital micromirror device (DMD). The fluorescent emission is projected onto the DMD to facilitate the sequential segmentation of fluorescence from individual objects in the field-of-view, allowing for both image acquisition and fluorescence lifetime determination of the assay. The sensitivity of PLI is demonstrated by manually segmenting fluorescence from fixed cell assays. We also demonstrate an automated implementation of PLI, using a camera as a feedback mechanism to segment fluorescence produced by emitting objects of interest in the imaging field-of-view, highlighting the advantages of measurement only in areas where valuable information exists. As a result, PLI is able to reduce acquisition time of fluorescence lifetime data by at least an order of magnitude compared to laser scanning implementations.

Video-rate volumetric neuronal imaging using 3D targeted illumination

Fast volumetric microscopy is required to monitor large-scale neural ensembles with high spatio-temporal resolution. Widefield fluorescence microscopy can image large 2D fields of view at high resolution and speed while remaining simple and costeffective. A focal sweep add-on can further extend the capacity of widefield microscopy by enabling extended-depth-of-field (EDOF) imaging, but suffers from an inability to reject out-of-focus fluorescence background. Here, by using a digital micromirror device to target only in-focus sample features, we perform EDOF imaging with greatly enhanced contrast and signal-to-noise ratio, while reducing the light dosage delivered to the sample. Image quality is further improved by the application of a robust deconvolution algorithm. We demonstrate the advantages of our technique for in vivo calcium imaging in the mouse brain.

Neuronal imaging with ultrahigh dynamic range multiphoton microscopy

Multiphoton microscopes are hampered by limited dynamic range, preventing weak sample features from being detected in the presence of strong features, or preventing the capture of unpredictable bursts in sample strength. We present a digital electronic add-on technique that vastly improves the dynamic range of a multiphoton microscope while limiting potential photodamage. The add-on provides real-time negative feedback to regulate the laser power delivered to the sample, and a log representation of the sample strength to accommodate ultrahigh dynamic range without loss of information. No microscope hardware modifications are required, making the technique readily compatible with commercial instruments. Benefits are shown in both structural and in-vivo functional mouse brain imaging applications.

Spatially-controlled illumination microscopy

Live-cell and live-tissue imaging using fluorescence optical microscopes presents an inherent trade-off between image quality and photodamage. Spatially-controlled illumination microscopy (SCIM) aims to strike the right balance between obtaining good image quality and minimizing the risk of photodamage. In traditional imaging, illumination is performed with a spatially-uniform light dose resulting in spatially-variable detected signals. SCIM adopts an alternative imaging approach where illumination is performed with a spatially-variable light dose resulting in spatially-uniform detected signals. The actual image information of the biological specimen in SCIM is predominantly encoded in the illumination profile. SCIM uses real-time spatial control of illumination in the imaging of fluorescent biological specimens. This alternative imaging paradigm reduces the overall illumination light dose during imaging, which facilitates prolonged imaging of live biological specimens by minimizing photodamage without compromising image quality. Additionally, the dynamic range of a SCIM image is no longer limited by the dynamic range of the detector (or camera), since it employs a uniform detection strategy. The large dynamic range of SCIM is predominantly determined by the illumination profile, and is advantageous for imaging both live and fixed biological specimens. In the present review, the concept and working mechanisms of SCIM are discussed, together with its application in various types of optical microscopes.

Adaptive illumination reduces photobleaching in structured illumination microscopy

Photobleaching is a major factor limiting the observation time in fluorescence microscopy. We achieve photobleaching reduction in structured illumination microscopy (SIM) by locally adjusting the illumination intensities according to the sample. Adaptive SIM is enabled by a digital micro-mirror device (DMD), which provides a projection of the grayscale illumination patterns. We demonstrate a reduction in photobleaching by a factor of three in adaptive SIM compared to the non-adaptive SIM based on a spot grid scanning approach. Our proof-of-principle experiments show great potential for DMD-based microscopes to become a more useful tool in live-cell SIM imaging.

Development of new photon-counting detectors for single-molecule fluorescence microscopy

Two optical configurations are commonly used in single-molecule fluorescence microscopy: point-like excitation and detection to study freely diffusing molecules, and wide field illumination and detection to study surface immobilized or slowly diffusing molecules. Both approaches have common features, but also differ in significant aspects. In particular, they use different detectors, which share some requirements but also have major technical differences. Currently, two types of detectors best fulfil the needs of each approach: single-photon-counting avalanche diodes (SPADs) for point-like detection, and electron-multiplying charge-coupled devices (EMCCDs) for wide field detection. However, there is room for improvements in both cases. The first configuration suffers from low throughput owing to the analysis of data from a single location. The second, on the other hand, is limited to relatively low frame rates and loses the benefit of single-photon-counting approaches. During the past few years, new developments in point-like and wide field detectors have started addressing some of these issues. Here, we describe our recent progresses towards increasing the throughput of single-molecule fluorescence spectroscopy in solution using parallel arrays of SPADs. We also discuss our development of large area photon-counting cameras achieving subnanosecond resolution for fluorescence lifetime imaging applications at the single-molecule level.

Live-cell migration and adhesion turnover assays

Fluorescence microscopy has revolutionized the way live-cell imaging is achieved. At the same time, it is also potentially harmful to a living specimen. Therefore, the specimen must be monitored for viability and health before, during, and after imaging sessions. Methods for monitoring cell viability and health will be discussed in this chapter. Another key to successful live-cell imaging is to minimize light exposure as much as possible. A summary of strategies for minimizing light exposure including maximizing the light throughput of the microscope and the sensitivity of light detection is presented. Various fluorescence microscopy techniques are presented with a focus on how the light is delivered to the sample (i.e., light density) and pros and cons for use with living specimens. The reader is also directed to other publications that go into these topics in more detail. Methods are described on how to prepare samples for single cell migration assays, how to measure cell migration rates (e.g., bright-field, semi-automated, and automated), and how to measure focal adhesion turnover rates. Details of how to correct images for background intensity and field-illumination uniformity artifacts for quantitative imaging are also described. Overall, this chapter will be helpful to scientists who are interested in imaging live specimens using fluorescence microscopy techniques. It will be of particular interest to anyone wanting to perform quantitative fluorescence imaging, and wanting to measure cell migration rates, and focal adhesion dynamics.

flatFLIM: enhancing the dynamic range of frequency domain FLIM

Fluorescence Lifetime Imaging Microscopy (FLIM) is a quantitative technique to probe the nanoenvironment of fluorescent molecules. It is the most robust way to quantify Förster Resonance Energy Transfer (FRET) as it allows reliable differentiation between concentration changes and quenching. In this way, molecular interactions can be imaged in single living cells. The most common wide-field implementation is homodyne Frequency Domain (FD) FLIM, which determines the fluorescence lifetime by measuring the phase and modulation changes of the fluorescence in each pixel upon excitation with a light source modulated at a high frequency. The fluorescence lifetimes are derived from a stack of images acquired at different phase shifts between excitation and detection. In this work we describe a simple method to enhance the dynamic range of FD-FLIM based on precompensating the expected fluorescence modulation by varying the laser power through the phase stack. We show theoretically and experimentally that most of the dynamic range of the camera can be recovered to quantify cells with different intensities. This improvement can be added to any FD-FLIM setup with minimal modifications, enhancing the throughput of information content.

Predictive-focus illumination for reducing photodamage in live-cell microscopy

Due to photobleaching and phototoxicity induced by high-intensity excitation light, the number of fluorescence images that can be obtained in live cells is always limited. This limitation becomes particularly prominent in multidimensional recordings when multiple Z-planes are captured at every time point. Here we present a simple technique, termed predictive-focus illumination (PFI), which helps to minimize cells' exposure to light by decreasing the number of Z-planes that need to be captured in live-cell 3D time-lapse recordings. PFI utilizes computer tracking to predict positions of objects of interest (OOIs) and restricts image acquisition to small dynamic Z-regions centred on each OOI. Importantly, PFI does not require hardware modifications and it can be easily implemented on standard wide-field and spinning-disc confocal microscopes.