Automated Holotomographic Microscopy
Figure 6: Quantitative study of the impact of phototoxicity on mammalian cell growth dynamics using label-free holotomographic microscopy. (a) Cell segmentations are displayed in multicolor overlay. Control 3T3-derived pre-adipocytes (yellow frame), treated with MitoTracker in the absence (green frame) or presence of fluorescence excitation (excitation power 1% [blue frame] or 5% [brown frame]) show either perfect growth (yellow frame), or apoptosis after 10h30m (green frame), 7h30m (blue frame), and 2h30m (brown frame) of exposure to a CoolLED light source. (b) Regular plotting of the average and standard deviation of single-cell area, compactness, and dry mass for the control 3T3-derived pre-adipocytes (control, yellow line), or treated with MitoTracker, without fluorescence excitation (MitoTracker only, green line) or with fluores- cence excitation (excitation power 1% [blue line] or 5% [brown line]). (c) Refractive index image of a cell undergoing a typical apoptotic cell death found in the condition MitoTracker + excitation power 1%. (d) Refractive index image of a cell undergoing a typical necrotic cell death found in the condition MitoTracker + excitation power 5%.
does not involve massive leakage of material from the cells. However, at 5% power cells shrink much faster and their dry mass decreases quickly, indicating a loss of membrane integrity. Altogether, these dynamics indicate two types of cell death, one more controlled and lengthy, corresponding to typical apopto- sis (Figure 6c), while the other is sudden and destructive, cor- responding more to necroptosis (Figure 6d) [33].
Conclusion Te CX-A device is the first system to automate HTM
microscopy. Moreover, its new soſtware environment allows for unique quantifications using a proprietary label-free cell
2021 September •
www.microscopy-today.com
segmentation environment called EA. EA uses AI-aided sig- nal analysis and advanced object detection techniques to create a simple but powerful quantification pipeline. Tus, users of EA have access to a set of advanced algorithms without a long setup time. Finally, EA is remarkable by its existence in a global context where similar performances would only be provided by heavy deep-learning (DL) approaches. DL techniques come with major limitations: absence of globally applicable models, high demands on training and validation of data, and vast possibilities for setups, which effectively make DL and more generally AI-aided approaches impossible to generalize, keep- ing them away from widespread use [34]. With its capacity to
31
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68 |
Page 69 |
Page 70 |
Page 71 |
Page 72 |
Page 73 |
Page 74 |
Page 75 |
Page 76 |
Page 77 |
Page 78 |
Page 79 |
Page 80
