Semi-blind sparse affine spectral unmixing of autofluorescence-contaminated micrographs

A Framework of Multi-View Machine Learning for Biological Spectral Unmixing of Fluorophores with Overlapping Excitation and Emission Spectra

The accuracy in assigning fluorophore identity and abundance, termed spectral unmixing, in biological fluorescence microscopy images remains challenging due to the unavoidable and significant overlap in emission spectra among fluorophores. In conventional laser scanning confocal spectral microscopy, fluorophore information is acquired by recording emission spectra with a single combination of discrete excitation wavelengths. As a matter of fact, organic fluorophores have not only unique emission spectral signatures but also have unique and characteristic excitation spectra. In this paper, we propose a generalized multi-view machine learning approach, which makes use of both excitation and emission spectra to greatly improve the accuracy in differentiating multiple highly overlapping fluorophores in a single image. By recording emission spectra of the same field with multiple combinations of excitation wavelengths, we obtain data representing these different views of the underlying fluorophore distribution in the sample. We then propose a framework of multi-view machine learning methods, which allows us to flexibly incorporate noise information and abundance constraints, to extract the spectral signatures of fluorophores from their reference images and to efficiently recover their corresponding abundances in unknown mixed images. Numerical experiments on simulated image data demonstrate the method's efficacy in improving accuracy, allowing for the discrimination of 100 fluorophores with highly overlapping spectra. Furthermore, validation on images of mixtures of fluorescently labeled E. coli demonstrates the power of the proposed multi-view strategy in discriminating fluorophores with spectral overlap in real biological images.

BINGO: a blind unmixing algorithm for ultra-multiplexing fluorescence images

MOTIVATION

Spectral imaging is often used to observe different objects with multiple fluorescent labels to reveal the development of the biological event. As the number of observed objects increases, the spectral overlap between fluorophores becomes more serious, and obtaining a "pure" picture of each fluorophore becomes a major challenge. Here, we propose a blind spectral unmixing algorithm called BINGO (Blind unmixing via SVD-based Initialization Nmf with project Gradient descent and spare cOnstrain), which can extract all kinds of fluorophores more accurately from highly overlapping multichannel data, even if the spectra of the fluorophores are extremely similar or their fluorescence intensity varies greatly.

RESULTS

BINGO can isolate up to 10 fluorophores from spectral imaging data for a single excitation. nine-color living HeLa cells were visualized distinctly with BINGO. It provides an important algorithmic tool for multiplex imaging studies, especially in intravital imaging. BINGO shows great potential in multicolor imaging for biomedical sciences.

AVAILABILITY AND IMPLEMENTATION

The source code used for this paper is available with the test data at https://github.com/Xinyuan555/BINGO_unmixing.

Comparing Performance of Spectral Image Analysis Approaches for Detection of Cellular Signals in Time-Lapse Hyperspectral Imaging Fluorescence Excitation-Scanning Microscopy

Hyperspectral imaging (HSI) technology has been applied in a range of fields for target detection and mixture analysis. While HSI was originally developed for remote sensing applications, modern uses include agriculture, historical document authentication, and medicine. HSI has also shown great utility in fluorescence microscopy. However, traditional fluorescence microscopy HSI systems have suffered from limited signal strength due to the need to filter or disperse the emitted light across many spectral bands. We have previously demonstrated that sampling the fluorescence excitation spectrum may provide an alternative approach with improved signal strength. Here, we report on the use of excitation-scanning HSI for dynamic cell signaling studies-in this case, the study of the second messenger Ca2+. Time-lapse excitation-scanning HSI data of Ca2+ signals in human airway smooth muscle cells (HASMCs) were acquired and analyzed using four spectral analysis algorithms: linear unmixing (LU), spectral angle mapper (SAM), constrained energy minimization (CEM), and matched filter (MF), and the performances were compared. Results indicate that LU and MF provided similar linear responses to increasing Ca2+ and could both be effectively used for excitation-scanning HSI. A theoretical sensitivity framework was used to enable the filtering of analyzed images to reject pixels with signals below a minimum detectable limit. The results indicated that subtle kinetic features might be revealed through pixel filtering. Overall, the results suggest that excitation-scanning HSI can be employed for kinetic measurements of cell signals or other dynamic cellular events and that the selection of an appropriate analysis algorithm and pixel filtering may aid in the extraction of quantitative signal traces. These approaches may be especially helpful for cases where the signal of interest is masked by strong cellular autofluorescence or other competing signals.

Unmixing Biological Fluorescence Image Data with Sparse and Low-Rank Poisson Regression

Multispectral biological fluorescence microscopy has enabled the identification of multiple targets in complex samples. The accuracy in the unmixing result degrades (1) as the number of fluorophores used in any experiment increases and (2) as the signal-to-noise ratio in the recorded images decreases. Further, the availability of prior knowledge regarding the expected spatial distributions of fluorophores in images of labeled cells provides an opportunity to improve the accuracy of fluorophore identification and abundance. We propose a regularized sparse and low-rank Poisson unmixing approach (SL-PRU) to deconvolve spectral images labeled with highly overlapping fluorophores which are recorded in low signal-to-noise regimes. Firstly, SL-PRU implements multi-penalty terms when pursuing sparseness and spatial correlation of the resulting abundances in small neighborhoods simultaneously. Secondly, SL-PRU makes use of Poisson regression for unmixing instead of least squares regression to better estimate photon abundance. Thirdly, we propose a method to tune the SL-PRU parameters involved in the unmixing procedure in the absence of knowledge of the ground truth abundance information in a recorded image. By validating on simulated and real-world images, we show that our proposed method leads to improved accuracy in unmixing fluorophores with highly overlapping spectra.

Development of Spectral Imaging Cytometry

Spectral flow cytometry is a new technology that enables measurements of fluorescent spectra and light scattering properties in diverse cellular populations with high precision. Modern instruments allow simultaneous determination of up to 40+ fluorescent dyes with heavily overlapping emission spectra, discrimination of autofluorescent signals in the stained specimens, and detailed analysis of diverse autofluorescence of different cells-from mammalian to chlorophyll-containing cells like cyanobacteria. In this paper, we review the history, compare modern conventional and spectral flow cytometers, and discuss several applications of spectral flow cytometry.

Live-cell fluorescence spectral imaging as a data science challenge

Live-cell fluorescence spectral imaging is an evolving modality of microscopy that uses specific properties of fluorophores, such as excitation or emission spectra, to detect multiple molecules and structures in intact cells. The main challenge of analyzing live-cell fluorescence spectral imaging data is the precise quantification of fluorescent molecules despite the weak signals and high noise found when imaging living cells under non-phototoxic conditions. Beyond the optimization of fluorophores and microscopy setups, quantifying multiple fluorophores requires algorithms that separate or unmix the contributions of the numerous fluorescent signals recorded at the single pixel level. This review aims to provide both the experimental scientist and the data analyst with a straightforward description of the evolution of spectral unmixing algorithms for fluorescence live-cell imaging. We show how the initial systems of linear equations used to determine the concentration of fluorophores in a pixel progressively evolved into matrix factorization, clustering, and deep learning approaches. We outline potential future trends on combining fluorescence spectral imaging with label-free detection methods, fluorescence lifetime imaging, and deep learning image analysis.

When light meets biology - how the specimen affects quantitative microscopy

Fluorescence microscopy images should not be treated as perfect representations of biology. Many factors within the biospecimen itself can drastically affect quantitative microscopy data. Whereas some sample-specific considerations, such as photobleaching and autofluorescence, are more commonly discussed, a holistic discussion of sample-related issues (which includes less-routine topics such as quenching, scattering and biological anisotropy) is required to appropriately guide life scientists through the subtleties inherent to bioimaging. Here, we consider how the interplay between light and a sample can cause common experimental pitfalls and unanticipated errors when drawing biological conclusions. Although some of these discrepancies can be minimized or controlled for, others require more pragmatic considerations when interpreting image data. Ultimately, the power lies in the hands of the experimenter. The goal of this Review is therefore to survey how biological samples can skew quantification and interpretation of microscopy data. Furthermore, we offer a perspective on how to manage many of these potential pitfalls.

Visualizing the Invisible: Advanced Optical Microscopy as a Tool to Measure Biomechanical Forces

The importance of mechanical force in biology is evident across diverse length scales, ranging from tissue morphogenesis during embryo development to mechanotransduction across single adhesion proteins at the cell surface. Consequently, many force measurement techniques rely on optical microscopy to measure forces being applied by cells on their environment, to visualize specimen deformations due to external forces, or even to directly apply a physical perturbation to the sample via photoablation or optogenetic tools. Recent developments in advanced microscopy offer improved approaches to enhance spatiotemporal resolution, imaging depth, and sample viability. These advances can be coupled with already existing force measurement methods to improve sensitivity, duration and speed, amongst other parameters. However, gaining access to advanced microscopy instrumentation and the expertise necessary to extract meaningful insights from these techniques is an unavoidable hurdle. In this Live Cell Imaging special issue Review, we survey common microscopy-based force measurement techniques and examine how they can be bolstered by emerging microscopy methods. We further explore challenges related to the accompanying data analysis in biomechanical studies and discuss the various resources available to tackle the global issue of technology dissemination, an important avenue for biologists to gain access to pre-commercial instruments that can be leveraged for biomechanical studies.