15. Optionally, detach the camera from the microscope; immediately cap the camera sensor and the microscope camera port to
prevent dust ingress.
16. Optionally, disconnect the BCS196 stage body from the
microscope stand, and remove the mounting hardware from
the stage.
3.9 Data Analysis
To analyze the kinetics of the mechanisms that cause intracellular
ice crystallization, the cumulative hazard of intracellular ice formation can be computed using the Nelson-Aalen estimator, a versatile
nonparametric method that can account for competition between
multiple mechanisms, as well as data censoring [12, 26]. Nonetheless, in the procedure described below, no attempt is made to
distinguish between different mechanisms of intracellular ice formation (i.e., only the total rate of intracellular ice formation is
determined), and data censoring is treated only in a rudimentary
manner. It is assumed that one or more replicate cryomicroscopy
experiments were performed and that high-speed video recordings
were acquired starting from the beginning of the rapid-cooling
ramp (e.g., Ramp 4 in Table 1) and ending before the beginning
of the warming ramp (e.g., Ramp 5 in Table 1).
1. In every video, display the first post-trigger frame (see Note
100), and assign each cell a unique identifying label. Reviewing
the full length of each video (starting from the first post-trigger
frame), if any of these cells becomes obscured at any point
during the experiment (e.g., by drifting out of the field of
view, or out of focus), omit such cells from further analysis
(see Note 101).
2. For each remaining cell, using slow-motion and frame-byframe playback, find the video image in which intracellular ice
crystallization is initiated (i.e., the first frame in which any
evidence of intracellular ice can be seen) (see Notes 102 and
103). Record the timestamp of each such video frame (with
time measured relative to the time at which the camera
detected the falling edge of the trigger signal).
3. For each video, also go to the first post-trigger frame and to the
final video frame, and record the corresponding timestamps
(with time measured relative to the time at which the camera
detected the falling edge of the trigger signal).
4. Using the Linksys32 software, convert each saved temperature
data file from its native format (.iml) to a text file with commaseparated values (see Note 104), then open the resulting file
using any standard spreadsheet or word processing software, or
optionally, import the data into a computational software platform such as MATLAB (see Note 14).
240
Jens O. M. Karlsson
prevent dust ingress.
16. Optionally, disconnect the BCS196 stage body from the
microscope stand, and remove the mounting hardware from
the stage.
3.9 Data Analysis
To analyze the kinetics of the mechanisms that cause intracellular
ice crystallization, the cumulative hazard of intracellular ice formation can be computed using the Nelson-Aalen estimator, a versatile
nonparametric method that can account for competition between
multiple mechanisms, as well as data censoring [12, 26]. Nonetheless, in the procedure described below, no attempt is made to
distinguish between different mechanisms of intracellular ice formation (i.e., only the total rate of intracellular ice formation is
determined), and data censoring is treated only in a rudimentary
manner. It is assumed that one or more replicate cryomicroscopy
experiments were performed and that high-speed video recordings
were acquired starting from the beginning of the rapid-cooling
ramp (e.g., Ramp 4 in Table 1) and ending before the beginning
of the warming ramp (e.g., Ramp 5 in Table 1).
1. In every video, display the first post-trigger frame (see Note
100), and assign each cell a unique identifying label. Reviewing
the full length of each video (starting from the first post-trigger
frame), if any of these cells becomes obscured at any point
during the experiment (e.g., by drifting out of the field of
view, or out of focus), omit such cells from further analysis
(see Note 101).
2. For each remaining cell, using slow-motion and frame-byframe playback, find the video image in which intracellular ice
crystallization is initiated (i.e., the first frame in which any
evidence of intracellular ice can be seen) (see Notes 102 and
103). Record the timestamp of each such video frame (with
time measured relative to the time at which the camera
detected the falling edge of the trigger signal).
3. For each video, also go to the first post-trigger frame and to the
final video frame, and record the corresponding timestamps
(with time measured relative to the time at which the camera
detected the falling edge of the trigger signal).
4. Using the Linksys32 software, convert each saved temperature
data file from its native format (.iml) to a text file with commaseparated values (see Note 104), then open the resulting file
using any standard spreadsheet or word processing software, or
optionally, import the data into a computational software platform such as MATLAB (see Note 14).
240
Jens O. M. Karlsson