Ghost imaging is a technique used to produce an object's image without using a spatially resolving detector. Here we develop a technique we term "ghost cytometry," an image-free ultrafast fluorescence "imaging" cytometry based on a single-pixel detector. Spatial information obtained from the motion of cells relative to a static randomly patterned optical structure is compressively converted into signals that arrive sequentially at a single-pixel detector. Combinatorial use of the temporal waveform with the intensity distribution of the random pattern allows us to computationally reconstruct cell morphology. More importantly, we show that applying machine-learning methods directly on the compressed waveforms without image reconstruction enables efficient image-free morphology-based cytometry.
Flow cytometers measure the optical properties of particles to classify microbes. Recent innovations have allowed oceanographers to collect flow cytometry data continuously during research cruises, leading to an explosion of data and new challenges for the classification task.The massive scale, time-varying underlying populations, and noisy measurements motivate the development of new classification methods. We describe the problem, the data, and some preliminary results demonstratingthe difficulty with conventional methods.
In the near future, clinical trials will use machine learning routinely to analyze data. Today, however, machine learning is still at an adoption midpoint--no longer pioneering, but not yet ubiquitous. For example, David Craford, president and chief executive officer, Cytobank (now a Beckman Coulter Life Sciences company), shares that, "In a clinical trials data analysis session at the CYTO 2019 Conference on flow cytometry in June, a live, 104-person survey indicated that 53% are using'an unsupervised approach such as FlowSOM' (a machine learning algorithm) to analyze and visualize cytometry data sets." These numbers are sure to grow in the immediate future as the trends in increasing size and quantities of datasets, coupled with decreasing costs to generate them, continue. Machine learning--self-training algorithms that sift through massive amounts of data looking for unknown signals in the noise--has emerged as an absolute necessity for making sense of all of that data.
Data used in Flow Cytometry present pronounced variability due to biological and technical reasons. Biological variability is a well known phenomenon produced by measurements on different individuals, with different characteristics such as age, sex, etc... The use of different settings for measurement, the variation of the conditions during experiments or the different types of flow cytometers are some of the technical sources of variability. This high variability makes difficult the use of supervised machine learning for identification of cell populations. We propose optimalFlowTemplates, based on a similarity distance and Wasserstein barycenters, which clusterizes cytometries and produces prototype cytometries for the different groups. We show that supervised learning restricted to the new groups performs better than the same techniques applied to the whole collection. We also present optimalFlowClassification, which uses a database of gated cytometries and optimalFlowTemplates to assign cell types to a new cytometry. We show that this procedure can outperform state of the art techniques in the proposed datasets. Our code and data are freely available as R packages at https://github.com/HristoInouzhe/optimalFlow and https://github.com/HristoInouzhe/optimalFlowData.
Extracellular vesicles (EVs) such as exosomes (70 nm – 160 nm in diameter) and microvesicles (100 nm – 1,000 nm diameter) can be harvested from cell-culture supernatants and from all bodily fluids. Current standard techniques to visualize, quantify, and characterize EVs are electron microscopy, nanoparticle tracking analyses, and dynamic light scattering. To further characterize and discriminate EVs, more exact high-throughput technologies to analyze their surface are highly desired. Although conventional flow cytometry is limited to measuring particles down to approximately 300 nm – 500 nm, a relatively new flow-cytometric method--called "imaging flow cytometry"--allows for the analysis of EVs smaller than 300 nm. This webinar will introduce viewers to the challenges, limitations, and pitfalls of flow cytometry-based EV analysis, and to the imaging flow cytometry methodology.