PML_buddy.py

# This file is ment to be placed in the /plugins/ directory of FIJI and is ment
# to be run from within the plugins menu.
#
# Copyright 2016 Jens Eriksson 

# Permission is hereby granted, free of charge, to any person obtaining a 
# copy of this software and associated documentation files (the 
# "Software"), to deal in the Software without restriction, including 
# without limitation the rights to use, copy, modify, merge, publish, 
# distribute, sublicense, and/or sell copies of the Software, and to 
# permit persons to whom the Software is furnished to do so, subject to 
# the following conditions: 

# The above copyright notice and this permission notice shall be included 
# in all copies or substantial portions of the Software. 

# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS 
# OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF 
# MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. 
# IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY 
# CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, 
# TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE 
# SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. 


import ij.gui.Roi as Roi
import ij.gui.OvalRoi as OvalRoi
import ij.gui.Overlay as Overlay
from ij.measure import ResultsTable
from ij import WindowManager as WindowManager
from ij.plugin.frame import RoiManager as RoiManager
from ij import IJ, ImagePlus, ImageStack
import ij.process.ImageStatistics as ImageStatistics
from ij.measure import Measurements as Measurements
from ij import IJ as IJ
import java.awt.Color as Color
from ij.gui import Plot as Plot
from ij.gui import PlotWindow as PlotWindow
from ij.gui import GenericDialog
import math
from ij.measure import CurveFitter as CurveFitter



def setupDialog(imp):
    """
    Creates a GenericDialog object where the relevant settings for the plugin
    can be accessed and set through a convenient GUI

    Args:
        imp: ij.ImagePlus object, usually the currently active window.

    Returns:
        A GenericDialog Object containig all the desired settings for the
        analysis
    """

    gd = GenericDialog("PML Buddy options")
    gd.addMessage("Welcome to PML Buddy, you are analyzing: "+imp.getTitle())
    calibration = imp.getCalibration()

    if calibration.frameInterval > 0:
        default_interval = calibration.frameInterval
        default_timeunit = calibration.getTimeUnit()
    else:
        default_interval = 0
        default_timeunit = None
        
    gd.addNumericField("Frame interval:", default_interval, 2)  #show 2 decimals    
    gd.addStringField("time unit", default_timeunit, 3)
    gd.addMessage("(optional)") #providing a unit is optional
    
    #Makes a list with the names of the channels in the imp
    
    channels = [str(ch) for ch in range(1, imp.getNChannels()+1)]
    
    #defaults to currnet active channel
    
    gd.addChoice("Channel to track:", channels, channels[imp.getChannel()-1]) 

    roichoises = ['Currently active Roi','First Roi in RoiManager']    
    gd.addChoice('Roi to use for tracking:', roichoises, roichoises[0])
    
    gd.addNumericField("Nr. of Roi centerings to do per frame:", 6, 0)
    gd.addNumericField("Diameter of analysis Roi (in pixels):)",20,0)
    gd.addSlider("Start tracking at frame:", 1, imp.getNFrames(),
                 imp.getFrame())
    gd.addSlider("Stop tracking at frame:", 1, imp.getNFrames(),
                 imp.getNFrames())
    
    gd.addCheckbox("Display tracking in new window", True)
    gd.addCheckbox("Show intensity plot", False)
    gd.addCheckbox("Show results table", False)
    gd.addCheckbox("Show cropped region", False)
    gd.addCheckbox("Use scaled analysis ROI", False)
    gd.addCheckbox("Do colocalization analysis on Ch1 & Ch2", True)
    gd.addCheckbox("Plot Colocalization coefficients", False)
    
    gd.showDialog()  
      
    if gd.wasCanceled(): #TODO proper exception handling
        IJ.log("User canceled dialog!")
        raise Exception("User canceled dialog!")
        return

    return gd

def roiCenterer(ip, roi, cal):
    """
    Centers the given roi on the center of mass inside the roi
    
    Args:
        ip: ImageProcessor
        roi: Region of intrest
        cal: Calibration of the ip
        
    Returns:
        an OvalRoi object which is centered on the center of mass of the input
          roi applied to the ImageProcessor
          
    """
    
    roi_w = roi.getFloatWidth()
    roi_h = roi.getFloatHeight()
    ip.setRoi(roi)
    stats = ImageStatistics.getStatistics(ip, ImageStatistics.CENTER_OF_MASS,
                                          cal)
    
    x = cal.getRawX(stats.xCenterOfMass)
    y = cal.getRawY(stats.yCenterOfMass)
    roi_x = x-roi_w/2
    roi_y = y-roi_h/2
    
    out_roi = OvalRoi(roi_x, roi_y, roi_w, roi_h)
    
    return out_roi

def roiScaler(roi, new_diameter):
    """
    Used if the tracking ROI has a hard time locking on to the object and
      needs therefore needs to be bigger than the analysis ROI.
    
    Agrs:
        roi: Region of intrest
        new_diameter: diameter of the returned roi
    
    Returns:
        OvalRoi centered on the input roi, but with a diameter of new_diameter
    """ 
    
    roi_x = roi.getXBase()
    roi_y = roi.getYBase()
    roi_w = roi.getFloatWidth()
    roi_h = roi.getFloatHeight()
    
    roi_x = roi_x + roi_w * 0.5 - new_diameter * 0.5
    roi_y = roi_y + roi_h * 0.5 - new_diameter * 0.5
    
    scaled_roi = OvalRoi(roi_x, roi_y, new_diameter, new_diameter)

    return scaled_roi
    
def channelStats(ip, channel, roi, resultdict, cal):
    """
    Records the mean intensity and X/Y positions of the center of mass for the
      roi in to the resultsDict.
    
    Args:
        ip: ImageProcessor
        channel: int channel number to analyze
        roi: roi to analyze
        resultdict: dict storing the results from the analysis, updated in place
        cal : calibration corresponding to the supplied ImageProcessor ip
    
    Returns:
        ImageProcessor, cropped from ip to the roi
    """
    ip.setRoi(roi)
    
    stats = ImageStatistics.getStatistics(ip, ImageStatistics.CENTER_OF_MASS,
                                          cal)
    x = cal.getRawX(stats.xCenterOfMass)
    y = cal.getRawY(stats.yCenterOfMass)
    
    resultdict['means_ch'+str(channel)].append(stats.mean)
    resultdict['ch'+str(channel)+'x'].append(x)
    resultdict['ch'+str(channel)+'y'].append(y)

    return ip.crop()

def colocRecorder(ip1, ip2, resultdict):
    """
    Args:
        ip1: ImageProcessor
        ip2: ImageProcessor
        resultdict: dict that stores the results
    
    Returns:
        nothing, updates resultdict
    """
    m = calcMandersCoefficients(ip1, ip2)
    resultdict['M1'].append(m[0])
    resultdict['M2'].append(m[1])
    resultdict['Pearson'].append(calcPearsonsCoefficient(ip1, ip2))
    resultdict['overlap_coefficient'].append(calcOverlapCoefficient(ip1, ip2))

    return
    
def calcOverlapCoefficient(ip1, ip2):
    """
    Calculates Manders Overlap Coeficcient, MOC, as
    specified in equation 2 in Manders et al. 1993. 
    Args:
        ip1, ip2: ImageProcessors of equal size
    
    Returns:
        float, representing the overlap coefficient
    """
    G = ip1.getPixels()
    R = ip2.getPixels()
    
    accum = 0
    Gsum = 0
    Rsum = 0

    for i in range(len(G)):
         accum += G[i]*R[i]
         Gsum += G[i]**2
         Rsum += R[i]**2

    if Gsum*Rsum==0:
        return 0
    
    return accum/math.sqrt(Gsum*Rsum)
    

def calcMandersCoefficients(ip1, ip2, th_G=0, th_R=0):
    """
    Calculates thresholded Mandlers colocalization coefficients, MCC.
    
    Thresholds defaults to 0, and as such calculates M1 & M2 as
    specified in Manders et al. (1993). If threshold values are supplied
    the function returns thresholded M1 & M2 values, as specified in
    Costes et al. (2004)
    
    Args:
        ip1, ip2: two imageProcessors of equal size
        th_G, th_R: threshold values for ip1 and ip2, defaults to 0
    
    Returns:
        floats M1, M2, representing Manders coefficients
    """
    ip1 = ip1.convertToFloatProcessor()
    ip2 = ip2.convertToFloatProcessor()
    G = ip1.getPixels()
    R = ip2.getPixels()

    
    Gcoloc = 0
    Rcoloc = 0
    
    for g, r in zip(G, R):
         if g > 0 and r > th_R:
             Rcoloc += r
         if r > 0 and g > th_G:
             Gcoloc += g
         
    Gsum = sum(G)
    Rsum = sum(R)
    
    if Gsum*Rsum==0:
        return 0,0
        
    return Gcoloc/float(Gsum), Rcoloc/float(Rsum)

def calcPearsonsCoefficient(ip1, ip2, Th_G=0, Th_R=0):
    """
    Calculates Pearson's correlation coeficcient, PCC. PCC describes the degree
    of overlap between two patterns. It provides information about the
    similarity of shape without regard to the average intensity of the signals.
    PCC varies from -1 to 1. Negative values are hard to interpret when degree
    of overlap is the quantity measured.
    
    
    Args:
        ip1, ip2: ImageProcessors of equal size
        Th1, Th2: Threshold values, calculates PCC for
        pixels above These values, defaluts to 0
        
    Returns:
        float R, representing Pearson's coefficient.
    """
    G = ip1.getPixels()
    R = ip2.getPixels()

    if (Th_G > 0) or (Th_R > 0):
        G,R = thresholder(G, R, Th_G, Th_R)

    Gsq = 0
    Rsq = 0
    
    Gavg=sum(G)/float(len(G)) #average pixel value in ip1
    Ravg=sum(R)/float(len(R)) #average pixel value in ip2

    num=0
    
    for i in range(len(G)):
         Rdiff = R[i]-Ravg
         Gdiff = G[i]-Gavg
         num+=Rdiff*Gdiff
         Gsq+=(Rdiff**2)
         Rsq+=(Gdiff**2)
         
    if Gsq*Rsq==0:
        return 0
    return num/(math.sqrt(Gsq*Rsq))

def thresholder(Ch1_pix, Ch2_pix, Th1, Th2):
    """Returns the pixels above thresholds 1 and 2.

    This function is called from inside the calcPearsonsCoefficient
    function if you supply it with at least one threshold value.

    Args:
        Ch1_pix: Array with the channel 1 pixels
        Ch2_pix: array with the channel 2 pixels
        Th1: Threshold for channel 1
        Th2: Threshold for channel 2
        
    Returns:
        A tuple containing the pixels that pass the threshold for 
        both channel 1 and channel 2.
        
        ([ch1],[ch2])
     """
    out1 = []
    out2 = []
        
    for g, r in zip(Ch1_pix, Ch2_pix):
        if (g > Th1) and (r > Th2):
            out1.append(g)
            out2.append(r)

    return out1, out2
    
def getLinfit(ch1_pix, ch2_pix): ##TODO implement this
    
    """
    Calculates the Costes threshold, T, as defined in Costes et al. (2004)
    The a and b constnants from the linear least square fit of Ch1 & Ch2 pixels
    is used to calculate T for ch2.
    """
    fitter = CurveFitter(ch1_pix, ch1_pix)
    fitter.doFit(CurveFitter.STRAIGHT_LINE)

    return fitter

    
# Start by getting the ImagePlus object in the active ImageJ window

imp = WindowManager.getCurrentImage()

# Store the Calibration object from the ImagePlus object in the variable cal

cal = imp.getCalibration()

# Run the setupDialog

gd = setupDialog(imp)  	

# Read out the options from the setupDialog

frame_interval = gd.getNextNumber()
time_unit = gd.getNextString()
channel_to_track = int(gd.getNextChoice())
roi_to_use = gd.getNextChoiceIndex() #0 for current, 1 for first in ROImanager
no_of_centerings = int(gd.getNextNumber())
analsis_roi_diameter = int(gd.getNextNumber())
start_frame = int(gd.getNextNumber())
stop_frame = int(gd.getNextNumber())

# Sanity check on start/stop frames

if (start_frame > stop_frame):
    IJ.showMessage("Start frame > Stop frame!")
    raise Exception("Start frame > Stop frame!")

no_frames_tracked=(stop_frame-start_frame)+1

# Setting flags

showTrackFlag=gd.getNextBoolean()
showPlotFlag=gd.getNextBoolean()
showResultsFlag=gd.getNextBoolean()
showCropFlag=gd.getNextBoolean()
analysisRoiFlag=gd.getNextBoolean()
colocalizationFlag=gd.getNextBoolean()
showColPlotFlag=gd.getNextBoolean()

# Set the frame interval in the calibration and store it back to the
# ImageProcessor

cal.frameInterval = frame_interval
cal.setTimeUnit(time_unit)
imp.setCalibration(cal)
    

# ImageStack class represents an expandable array of imgaes
stack = imp.getImageStack()

# Creates an empty stack with the same dimensions as imp
stack_track = imp.createEmptyStack() 

title = imp.getTitle()
n_channels = imp.getNChannels()
slice_to_track = 1 #Z-stack tracking is not implemented



# Get the chosen tracking ROI
if (roi_to_use == 0) and (imp.getRoi() is not None):
    roi_1 = imp.getRoi()

else: #TODO proper exception handling in case there are no ROIs
    roi_manager = RoiManager.getInstance()
    roi_list    = roi_manager.getRoisAsArray()
    roi_1 = roi_list[0]

# Get position and size of ROI
roi_x=roi_1.getXBase()
roi_y=roi_1.getYBase()
roi_w=roi_1.getFloatWidth()
roi_h=roi_1.getFloatHeight()

# In case we are analyzing a smaller region than we track
if analysisRoiFlag:
 
    stack_crop = ImageStack(int(analsis_roi_diameter),
                            int(analsis_roi_diameter))
else:
    stack_crop = ImageStack(int(roi_w), int(roi_h))


# A dictionary to keep track of results
result_dict={}
result_keys=['means_ch1','means_ch2','ch1x','ch1y','ch2x','ch2y']
    
if colocalizationFlag:
    extra_keys=['M1', 'M2', 'Pearson', 'overlap_coefficient']
    for extra in extra_keys: 
        result_keys.append(extra)
    
for key in result_keys:
    # each key holds a list of results
    result_dict[key]=[]
    

#loop through the frames that you want to track
for frame in range(start_frame, stop_frame+1):
    # Get the ImageProcessor of the channel to track at the current frame
    track_ip = stack.getProcessor(imp.getStackIndex(channel_to_track,
                                  slice_to_track, frame))
    
    track_roi = OvalRoi(roi_x, roi_y, roi_w, roi_h)
    
    #Do the Roi centering the desired number of times
    for i in range (no_of_centerings):
        track_roi=roiCenterer(track_ip, track_roi, cal)
    
    roi_x=track_roi.getXBase()
    roi_y=track_roi.getYBase()
    track_ip.setRoi(track_roi)

    if analysisRoiFlag:
        analysis_roi=roiScaler(track_roi, analsis_roi_diameter)
    else:
        analysis_roi=track_roi.clone()
    
    #Get Channel 1&2 IPs and apply the centered roi with the desired diameter
    
    ip1 = stack.getProcessor(imp.getStackIndex(1, slice_to_track, frame))
    ip1_crop=channelStats(ip1, 1, analysis_roi, result_dict, cal)
    
    ip2 = stack.getProcessor(imp.getStackIndex(2,slice_to_track,frame))
    ip2_crop=channelStats(ip2, 2, analysis_roi, result_dict, cal)

    if colocalizationFlag:
        colocRecorder(ip1_crop, ip2_crop, result_dict)

    if showTrackFlag:
        ip_track=track_ip.duplicate()
        ip_track.setColor(track_ip.maxValue())
        ip_track.draw(track_roi)
        
        if analysisRoiFlag:
            ip_track.setColor(track_ip.maxValue()/2)
            ip_track.draw(analysis_roi)
        stack_track.addSlice(ip_track)

    if showCropFlag:       
        #ip_crop.setColor(ip.maxValue())
        #ip_crop.draw(analysis_roi)
        stack_crop.addSlice(ip1_crop)
        stack_crop.addSlice(ip2_crop)


if showTrackFlag:    
    imp_track = ImagePlus(title+'_Processed', stack_track)
    imp_track.setCalibration(cal)
    imp_track.show()

if showCropFlag:
    imp_crop = IJ.createHyperStack(title+"_analysis_crop", int(roi_w),
                                   int(roi_h), n_channels, 1, no_frames_tracked,
                                   imp.getBitDepth())
    imp_crop.setStack(stack_crop)
    imp_crop.setCalibration(cal)
    imp_crop.show()

if showResultsFlag:
    IJ.run("Clear Results")
    rt=ResultsTable()
    for index in range(no_frames_tracked):
        rt.incrementCounter()
        for key in result_keys:
            rt.addValue(str(key),result_dict[key][index])
        
    rt.disableRowLabels()
    rt.show(title)

if showPlotFlag:
    maxc1 = max(result_dict['means_ch1'])
    maxc2 = max(result_dict['means_ch2'])
    
    if maxc1 > maxc2:
        plotlim = maxc1
    
    else:
        plotlim = maxc2

    if frame_interval > 0:
        time = [frame_interval*frame for frame in range(0,no_frames_tracked)]
        xlab="Time ("+time_unit+")"
    
    else:
        time=range(0,no_frames_tracked)
        xlab="frame"
    
    plot = Plot("Traced intensity curve for " + imp.getTitle(), xlab,
                "Mean intensity", [], [])
    
    plot.setLimits(1, max(time), 0, plotlim );
    plot.setLineWidth(2)
    
    plot.setColor(Color.GREEN)
    plot.addPoints(time, result_dict['means_ch1'], Plot.LINE)

    plot.setColor(Color.RED)
    plot.addPoints(time, result_dict['means_ch2'], Plot.LINE)
 
    plot.setColor(Color.black)
    plot.addLegend("Channel 1\nChannel 2")
    plot_window =  plot.show()

if showColPlotFlag:

    if frame_interval > 0:
        time = [frame_interval*frame for frame in range(0,no_frames_tracked)]
        xlab="Time ("+time_unit+")"
    else:
        time=range(0,no_frames_tracked)
        xlab="frame"

    plot = Plot("Correlationcoefficients over time for " + imp.getTitle(),
                xlab, "Correlation coefficient", [], [])
    
    plot.setLimits(1, max(time), -0.5, 3.2 );
    plot.setLineWidth(2)
    
    plot.setColor(Color.GREEN)
    plot.addPoints(time, result_dict['M1'], Plot.LINE)

    plot.setColor(Color.RED)
    plot.addPoints(time, result_dict['M2'], Plot.LINE)

    plot.setColor(Color.BLUE)
    plot.addPoints(time, result_dict['Pearson'], Plot.LINE)

    plot.setColor(Color.BLACK)
    plot.addPoints(time, result_dict['overlap_coefficient'], Plot.LINE)
    
    plot.addLegend("M1 (Ch1)\nM2 (Ch2)\nPearson\nOverlap coef")

    plot_window =  plot.show()