30. Use short exposure times and low laser power to minimize
photobleaching. Also, ensure that the camera is configured to
maximize the rate of image acquisition (Table 2). There is a
tradeoff between spatial resolution and temporal resolution
using SRRF: the more images in a SRRF sequence, the better
the reconstruction, but the longer the acquisition time and the
likelihood that cellular or molecular dynamics in the sample
will interfere with the reconstruction. In addition, longer
SRRF sequences will increase sample photobleaching. Typically, 50–100 images can be used to reconstruct a single
SRRF image from a live sample [24, 35]. Thus, single-plane
time-lapse SRRF can be performed with a temporal resolution
of under 4 s, if one accounts not only for the image acquisition
time, but also for the time necessary to save the images and
clear them from RAM. If acquiring Z-stacks, a temporal resolution of 1–2 min is possible.
31. If the name for an image acquisition already exists in the
selected directory, the new acquisition will be stored in a folder
with the same name and an incremented index at the end to
avoid overwriting the original data set. Each time point in the
SRRF acquisition will be saved in a different subfolder (time0
to timeN) within the acquisition folder. And each image in a
SRRF sequence will be automatically saved as an individual file,
using
the
following
naming
convention:
“img_XXXXXXXXX_Default0_YYY.tif”, where XXXXXXXXX
is the Z slice number, padded with zeros, and YYY is image
position within a specific SRRF sequence.
32. Default SRRF parameter values for radiality magnification, ring
radius, and axes can be used to start [24] (Fig. 5a, c). Radiality
magnification determines the number of subpixels that each
pixel is split into. For example, if this value is set to 5, each pixel
will be split into a 5 Â 5 grid of subpixels, and a 512 Â 512 raw
image sequence will lead to a 2560 Â 2560 super-resolution
image. Greater magnification values provide better final resolution, but increase the computation time necessary to reconstruct an image. The ring radius (in pixels) describes the local
area in which the convergence of the image gradient is quantified to determine radial symmetry and the presence of a fluorophore. A smaller ring can provide better resolution, but will
also increase noise of the reconstruction (Fig. 5c, f). Finally, the
axes value indicates the number of axes used to evaluate gradient convergence. A higher value (up to 8) is preferred for a
better assessment of radiality, but again, will increase the processing time during reconstruction.
33. The quality of SRRF images is affected by both acquisition
settings (number of images per burst, laser power, exposure
time) and image reconstruction ones (ring size, radiality
Live Imaging in Drosophila Embryos
219
photobleaching. Also, ensure that the camera is configured to
maximize the rate of image acquisition (Table 2). There is a
tradeoff between spatial resolution and temporal resolution
using SRRF: the more images in a SRRF sequence, the better
the reconstruction, but the longer the acquisition time and the
likelihood that cellular or molecular dynamics in the sample
will interfere with the reconstruction. In addition, longer
SRRF sequences will increase sample photobleaching. Typically, 50–100 images can be used to reconstruct a single
SRRF image from a live sample [24, 35]. Thus, single-plane
time-lapse SRRF can be performed with a temporal resolution
of under 4 s, if one accounts not only for the image acquisition
time, but also for the time necessary to save the images and
clear them from RAM. If acquiring Z-stacks, a temporal resolution of 1–2 min is possible.
31. If the name for an image acquisition already exists in the
selected directory, the new acquisition will be stored in a folder
with the same name and an incremented index at the end to
avoid overwriting the original data set. Each time point in the
SRRF acquisition will be saved in a different subfolder (time0
to timeN) within the acquisition folder. And each image in a
SRRF sequence will be automatically saved as an individual file,
using
the
following
naming
convention:
“img_XXXXXXXXX_Default0_YYY.tif”, where XXXXXXXXX
is the Z slice number, padded with zeros, and YYY is image
position within a specific SRRF sequence.
32. Default SRRF parameter values for radiality magnification, ring
radius, and axes can be used to start [24] (Fig. 5a, c). Radiality
magnification determines the number of subpixels that each
pixel is split into. For example, if this value is set to 5, each pixel
will be split into a 5 Â 5 grid of subpixels, and a 512 Â 512 raw
image sequence will lead to a 2560 Â 2560 super-resolution
image. Greater magnification values provide better final resolution, but increase the computation time necessary to reconstruct an image. The ring radius (in pixels) describes the local
area in which the convergence of the image gradient is quantified to determine radial symmetry and the presence of a fluorophore. A smaller ring can provide better resolution, but will
also increase noise of the reconstruction (Fig. 5c, f). Finally, the
axes value indicates the number of axes used to evaluate gradient convergence. A higher value (up to 8) is preferred for a
better assessment of radiality, but again, will increase the processing time during reconstruction.
33. The quality of SRRF images is affected by both acquisition
settings (number of images per burst, laser power, exposure
time) and image reconstruction ones (ring size, radiality
Live Imaging in Drosophila Embryos
219