Page 183 - IJB-8-1
P. 183

Yao, et al.
           be entangled with mask style as the output of this AD-GAN   procedure, two key steps, i.e., CLSM image collection and
           model,  that  is,  nuclei  segmentation  results.  Meanwhile,   nuclei segmentation are described as follows.
           spatial  structure  (the  shape  and  position  information  of
           nuclei) is preserved with content consistency during this   2.1. CLSM image collection from fibrous
           image-to-mask translation.                          scaffold-based cell culture
               The overall flow of this paper is organized as follows.
           The relevant methods are discussed in Section 2, including   CLSM  images  collected  from  three  types  of  fibrous
           CLSM  image  collection  in  fibrous  scaffold-based  cell   scaffolds  are  utilized  for  cell  segmentation  study:
           culture,  overview  of  AD-GAN,  its  training  strategy   Poly-E-caprolactone  (PCL)  scaffolds  and  two  types
           and  loss  function.  Section  3  covers  the  performance   of  surface  engineered  nanoporous  scaffolds,  namely,
                                                                                      [21]
           comparison of the segmentation methods including AD-  PCL-10-D  and  PCL-20-D .  The  schematic  diagram
           GAN, CycleGAN, Squassh, and CellProfiler under low/  of  scaffold  with  fiber  stacking  structure  is  shown  in
           high cell density or cross cell lines. Cell adhesion and   (Figure 3A), the scanning electron microscope (SEM)
           proliferation analysis based on the segmentation results   image  of  the  overall  fibrous  PCL  scaffold  structure  is
           are  reported  in  Section  4.  Finally,  the  conclusion  and   shown  in  (Figure  3B),  and  the  cross-section  of  fiber
           future work are set out in Section 5.               stacking  is  shown  in  (Figure  3C).  It  can  be  seen  that
                                                               the pores of the scaffold are uniformly distributed within
           2. Methods                                          the  structure,  and  fibers  are  well  oriented  and  orderly
           The  workflow  of  cell-scaffold  interaction  analysis   stacked  in  a  layer-by-layer  manner.  The  SEM  images
           through nuclei segmentation is shown in Figure 2. The   of  PCL-10-D  and  PCL-20-D  are  almost  the  same  as
           six-step  procedure  includes:  (A-F):  Scaffold-based  cell   those of PCL scaffolds. The overall thickness of these
           culture process; (B) CLSM image collection in the cell   12-layer scaffolds is between 110 µm and 127 µm, with
           culture  process;  (C)  AD-GAN  model  training;  (D)  3D   an average fiber diameter of around 10 µm. Meanwhile,
           nuclei segmentation results with position, size and shape   they have varied surface morphology. The fiber surface
           information;  (E)  cell-scaffold  interaction  analysis  using   of the PCL scaffolds is smooth (Figure 3D), the PCL-
           heat  map  produced  by  segmented  nuclei;  (F)  scaffold   10-D  scaffold  surface  is  covered  by  nanopits  with  an
           design revision. Figure 2(E) is obtained by projecting all   average size of  133.1 ± 47.4 nm (Figure 3E), and the
           the center point of segmented nuclei from step (D) on the   PCL-20-D  scaffold  surface  is  covered  by  much  larger
           XY  plane,  and  demonstrates  nuclei  density  distribution   nanopits and nanogrooves with lengths ranging from a
           using  the  brightness  of  red  color. To  further  clarify  the   few hundred nanometers to a few microns (Figure 3F).


           A                                  B                                     C
















            F                                 E                                     D












           Figure 2. (A-F) Schematic diagram of cell-scaffold interaction analysis using nuclei segmentation.

                                       International Journal of Bioprinting (2022)–Volume 8, Issue 1       169
   178   179   180   181   182   183   184   185   186   187   188