Table 1 iCLOTS comprises multiple workflows designed to address the needs of the widest range of experimental microscopy assays, with a focus on microfluidics
Workflow name | Adapted algorithm | Experimental device | Input data | Numerical result | Graphical result | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Single cell tracking applications (Fig. 3a–d, Supp. Figs. 3, 4) | |||||||||||||
Single cell tracking, option for channel flow or deformability | Crocker and Grier particle detection and linking | Commercial or custom microfluidics | Video Microscopy | Single-cell size and velocity with optional fluorescence intensity | Single-metric histograms, pairwise metric scatter plots | ||||||||
Velocity time course and profile | Shi-Tomasi corner detection and Kanade-Lucas-Tomasi feature tracking | Commercial or custom microfluidics | Video microscopy | Time course velocity, velocity profile | Scatter plot indicating mean, minimum, and maximum velocity values per frame and line graph showing channel-wise profile | ||||||||
Cell morphology (Fig. 5a, b) | Crocker and Grier particle detection | Static slide, commercial microfluidic, custom microfluidic, flow chamber | Single image(s) | Single-cell size and morphology, density | Single-metric histograms, pairwise metric scatter plots | ||||||||
Cell functionality (Fig. 5c, f) | Region analysis | Addl. fluorescence intensity | |||||||||||
Single-cell protrusion characterization (Fig. 5d) | Harris corner detection | Addl. protrusion count | |||||||||||
Transient adhesion time (Fig. 5e) | Crocker and Grier particle linking | Video microscopy | Addl. adhesion time | ||||||||||
Multiscale microfluidic accumulation applications (Fig. 6, Supp. Figs. 10, 11) | |||||||||||||
Surface (Fig. 6a) | Region analysis | Commercial or custom microfluidics | Single or time course image(s) | Multi-color channel occlusion and accumulation | Time course line graph of accumulation, occlusion over time | ||||||||
Complex geometry (Fig. 6b–d) | |||||||||||||
Microchannel (Fig. 6e, f) | Addl. values for multiple regions | Addl. representations of multiple regions | |||||||||||
Machine learning toolkit | |||||||||||||
Clustering (Fig. 3e, f, Fig. 5g, h, Fig. 6g–i, Supp. Fig. 2) | K-means clustering | iCLOTS or other numerical outputs | Cluster labels for individiual data points, frequency distribution per sample, goodness of clustering statistics | Correlation matrix, scree plot, scatter plot with clusters indicated, mosaic plot | |||||||||
