Figure 7: Classification of algal cells (Chlamydomonas reinhardtii) based on their lipid content by TS-QPI.

From: Deep Learning in Label-free Cell Classification

Figure 7

(a) Three-dimensional scatter plot based on size, protein concentration, and attenuation of the cells measured by TS-QPI, with 2D projections for every combination of two features. Inset: Conventional label-free flow cytometry using forward scattering and side scattering is not enough to distinguish the difference between high-lipid content and low-lipid content algal cells. TS-QPI is much more effective in separating the two algae populations. (b) ROC curves for binary classification of normal and lipid-rich algae species using ten-fold cross validation; blue curves show the classifier performance using all 16 biophysical features extracted from the TS-QPI quantitative images. Red, green, and orange curves show the classifier decision performance using only the best biophysical feature in each category: morphology (Diameter-RB in Table 1), optical phase (OPD-1 in Table 1), and optical loss (Absorption-2 in Table 1). The label-free selection of algal strains improves as more biophysical features are employed.