Flim

A collection of functions releavant to fluorescence lifetime imaging decay plotting and analysis.

Most used functions are the regionprops_omi and rectangular_to_phasor_point to compute g and s from a decay

Functions:

bin_image(image, bin_factor)	This function takes in an lifetime image and bins the decays of its histogram given a bin factor.
draw_universal_semicircle(...[, title, ...])	Draws the universal semicircle with the given frequency and returns the plt figure to draw over.
estimate_and_shift_irf(decay, irf_decay[, debug])	Estimates shift given a decay and an IRF.
ideal_sample_phasor(f, lifetime)	Generates a phasor for a specific lifetime at a given frequency.
lifetime_to_phasor(f, timebins, counts)	Time to frequency domain transformation
phasor_calculator(f, time, decays, IRF)	Given an array of decay(s) the rectangular g, s and phasors angle and magnitude will be computed and returned.
phasor_calibration(f, lifetime, timebins, counts)	Phasor Plot Calibration
phasor_to_rectangular(angle, magnitude)	Gets x and y position of phasor in rectangular coordinates
rectangular_to_phasor(g, s)	Takes an array(image) of g and s points and converts them to angle and magnitude phasor arrays
regionprops_omi(image_id, label_image[, ...])	Takes in labels image as well as nadh and fad images from SPCImage to return mean and stdev of each parameter per roi.

cell_analysis_tools.flim.bin_image(image, bin_factor)

This function takes in an lifetime image and bins the decays of its histogram given a bin factor. A bin of 2 will reduce the size of the image by 2. It will do so by forming a kernel of bin_factor x bin_factor and sum of those decays into a resulting pixel

Parameters:

• image (ndarray) – image to bin, must be a 3d array of shape (x,y,t)

• bin_factor (int) – pixel radius(including diagonals) of bins to create. Must be EVEN.

Returns:

binned_image – new image with binned pixels

Return type:

ndarray

cell_analysis_tools.flim.draw_universal_semicircle(laser_angular_frequency, title='Phasor Plot', suptitle='', debug=False)

Draws the universal semicircle with the given frequency and returns the plt figure to draw over.

Parameters:

• figure (matplotlib figure object) – figure object to draw semicircle over

• laser_angular_frequency (int) – rep rate or laser angular frequency, affects position of labeled points

Return type:

plt.figure object to plot your points over this figure

cell_analysis_tools.flim.estimate_and_shift_irf(decay, irf_decay, debug=False)

Estimates shift given a decay and an IRF.

Parameters:

• decay (ndarray) – 1D array containing decay curve

• irf_decay (ndarray) – 1D array containing IRF

• debug (bool) – Optional flag to display intermediate calculations

Returns:

• irf_decay_shifted (ndarray) – Shifted IRF as a 1d array

• shift (int) – value IRF was shifted by

Note

IRF and decay should NOT have low SNR or gradient function will produce incorrect alignment

cell_analysis_tools.flim.ideal_sample_phasor(f, lifetime)

Generates a phasor for a specific lifetime at a given frequency.

Parameters:

• f (int) – laser rep rate

• lifetime (float) – lifetime of desired single exponential sample

Returns:

• angle (float) – angle of ideal sample

• magnitude (float) – magnitude of ideal sample

>>print("Input is 80Mhz phasor at @ 2ns lifetime")
>>angle, magnitude = ideal_sample_phasor(f=80e6, lifetime=2e-9)
>>print(f"{angle=:.3f} rad | {magnitude=:.3f}")
angle=0.788 rad | magnitude=0.705

cell_analysis_tools.flim.lifetime_to_phasor(f, timebins, counts)

Time to frequency domain transformation

Parameters:

• f (int) – laser repetition angular frequency

• timebins (ndarray) – numpy array of timebins

• counts (ndarray) – photon counts of the histogram()

Returns:

• angle (float) – angle in radians

• magnitude (float) – magnitude of phasor

cell_analysis_tools.flim.phasor_calculator(f, time, decays, IRF)

Given an array of decay(s) the rectangular g, s and phasors angle and magnitude will be computed and returned.

Parameters:

• f (float) – Laser repetition rate.

• time (np.ndarray) – array of timebins.

• decays (np.ndarray) – Single decay or array of decays to compute phasor points for. Single decay should have the shape (x,),(x,t) and shape (x,y,t) for an array

• IRF (np.ndarray) – 1D array capturing irf decay.

Returns:

• m (float) – magnidue of phasor.

• phi (float) – angle of phasor.

• g (float) – g coordinate (x-axis).

• s (float) – s coordinate (y-axis).

Note

• You cannot compare two images directly due to them having different decay shift values between images, affeting g,s,m and phi locations

• For proper lifetime values, background subtraction is needed by taking ~ the last 1 or 1/2 ns timebins of decay)

cell_analysis_tools.flim.phasor_calibration(f, lifetime, timebins, counts)

Phasor Plot Calibration

Parameters:

• f (int) – laser repetition angular frequency

• lifetime (int) – lifetime of known sample in ns (single exponential decay)

• timebins (int) – timebins of samples

• counts (int) – photon counts of histogram

Returns:

• angle_offset {float} (difference in angle between known sample and actual)

• magnitude_offset {float} (difference in magnitude between known sample and actual)

Note

If no timebins or histograms passed then returns angle and phase of decay passed in.

cell_analysis_tools.flim.phasor_to_rectangular(angle, magnitude)

Gets x and y position of phasor in rectangular coordinates

Parameters:

• angle (float) – phasor angle in radians

• magnitude (float) – phasor magnitude

Returns:

• g (float) – x axis coordinate

• s (float) – y axis coordinate

cell_analysis_tools.flim.rectangular_to_phasor(g, s)

Takes an array(image) of g and s points and converts them to angle and magnitude phasor arrays

Parameters:

• g (float) – array of g coordinates (x-coordinates)

• s (float) – array of s coordinates (y-coordinates)

Returns:

PhasorArray object – lifetime_angles_array - array of angles for each pixel lifetime_magnitudes_array - array of magnitudes for each pixel

Return type:

float

cell_analysis_tools.flim.regionprops_omi(image_id: str, label_image: ndarray, im_nadh_intensity: ndarray | None = None, im_nadh_a1: ndarray | None = None, im_nadh_a2: ndarray | None = None, im_nadh_t1: ndarray | None = None, im_nadh_t2: ndarray | None = None, im_fad_intensity: ndarray | None = None, im_fad_a1: ndarray | None = None, im_fad_a2: ndarray | None = None, im_fad_t1: ndarray | None = None, im_fad_t2: ndarray | None = None, im_nadh_chi: ndarray | None = None, im_fad_chi: ndarray | None = None, other_props: list | None = None) → dict

Takes in labels image as well as nadh and fad images from SPCImage to return mean and stdev of each parameter per roi.

Parameters:

• label_image (ndarray) – labeled mask image.

• im_nadh_intensity (ndarray) – nadh intensity image.

• im_nadh_a1 (ndarray) – nadh alpha1 image.

• im_nadh_a2 (ndarray) – nadh alpha2 image .

• im_nadh_t1 (ndarray) – nadh tau 1 lifetime, short .

• im_nadh_t2 (ndarray) – nadh tau 2 lifetime, long.

• im_fad_intensity (ndarray) – nadh intensity image.

• im_fad_a1 (ndarray) – fad alpha 1 image.

• im_fad_a2 (ndarray) – fad alpha 2 image.

• im_fad_t1 (ndarray) – fad tau 1 lifetime, long.

• im_fad_t2 (ndarray) – fad tau 2 lifetime, short.

• other_props (list) – string list of additional parameters to compute on the binary mask see skimage regionprops for list of attributes

Note

See https://scikit-image.org/docs/dev/api/skimage.measure.html#skimage.measure.regionprops for a list of additiona properties you can compute

Returns:

dictionary of mean and standard deviations of omi parameters for each region.

Return type:

dict