Styjun

I study immune cells with the imaging software Imaris, which allows the masking of the surface of cells in a time-lapse experiment, and links each masked surface from one time frame to another, so that they can be tracked.
I work with 2D data sets and, for each time frame and each surface object (cell), I need to extract several 2D parameters (perimeter, major axis length, area), so that I can monitor cell shape variation over time. Unfortunately, Imaris gives several 3D parameters (volume of detected particles etc), but no 2D parameter.
Imaris is built so that data can be exported to Matlab, and I have been told it should be possible to use Matlab to extract these 2D parameters.

My objective is therefore to export my masked 2D surfaces (as a time-series) to Matlab, and extract the 2D parameters mentioned above.

With help from the Imaris support team I was able to correctly link Matlab and Imaris, import my data set, extract the mask data, and convert it to a binary matrix. With the provided code, this could only done for one time frame, however (out of 61). Still, I was able to convert this time frame (one image) to a binary matrix, and in turn the binary matrix to a binary image, from which I could extract the parameters I need (using the regionprops function).

My problem lies in doing this with the whole time series, which requires the original code to be modified. While I think I have made some progress, I believe I am stuck. Hence this question.

As you will see in the code below, I modified the "vindexT = 0" which selected only one time frame to "for vIndexT = 0:vSizeT" (where vSizeT is the maximum number of time frames, 61) which I believed would prompt Matlab to to do the same operation for all time frames. When I put the "end" operator just after the code extracting the mask data from Imaris, I do not get any error message. Which I hope means all the mask data had properly been imported (but I don't know how to verify that). However, when I run the next part of the code which should convert the mask data ("vMask") to a binary matrix, I get the error message I have pasted below.

Alternatively, I have tried to place the "end" operator after the code to convert the mask to a matrix (instead of having two loops I then only have one), but I got the same error message.

Here is the original code I was sent: this one allowed me to properly process one time frame:

%% %Get a Copy of the Dataset
vImarisDataSet = vImarisApplication.GetDataSet.Clone;
%%
% Get Surpass Surfaces Object - Please ensure that it is selected in the
% Surpass Tree
vImarisObject = vImarisApplication.GetSurpassSelection; 

%% %Convert the Object into the correct Data Type
vImarisObject = vImarisApplication.GetFactory.ToSurfaces(vImarisObject);

%% % Get the data dimensions
vDataMin = [vImarisDataSet.GetExtendMinX, vImarisDataSet.GetExtendMinY, vImarisDataSet.GetExtendMinZ];
vDataMax = [vImarisDataSet.GetExtendMaxX, vImarisDataSet.GetExtendMaxY, vImarisDataSet.GetExtendMaxZ];
vDataSize = [vImarisDataSet.GetSizeX, vImarisDataSet.GetSizeY, vImarisDataSet.GetSizeZ];

%%
% Create a new channel where the result will be sent
vNumberOfChannels = vImarisDataSet.GetSizeC;
vImarisDataSet.SetSizeC(vNumberOfChannels + 1);
vImarisDataSet.SetChannelName(vNumberOfChannels,['Surface Mask ', char(vImarisObject.GetName)]);
vImarisDataSet.SetChannelColorRGBA(vNumberOfChannels, 255*256*256+255*256+255);
%% %Get the Size of the Time Dimension
vDataSize = [vDataSize, vImarisDataSet.GetSizeT];
%% % Here only the first time point is considered to show the principle
vIndexT = 0;

%% %extract the Mask Data from Imaris
 vMaskDataSet = vImarisObject.GetMask( ...
 vDataMin(1), vDataMin(2), vDataMin(3), ...
 vDataMax(1), vDataMax(2), vDataMax(3), ...
 vDataSize(1), vDataSize(2), vDataSize(3), vIndexT);

%% %Convert the Imaris data into an Matlab Matrix
vMask = GetDataVolume(vMaskDataSet, 0, 0);

Here is what I have tried in order to import and convert the entire time series:

%%
% Create a new channel where the result will be sent
vNumberOfChannels = vImarisDataSet.GetSizeC;
vImarisDataSet.SetSizeC(vNumberOfChannels + 1);
vImarisDataSet.SetChannelName(vNumberOfChannels,['Surface Mask ', char(vImarisObject.GetName)]);
vImarisDataSet.SetChannelColorRGBA(vNumberOfChannels, 255*256*256+255*256+255);
aSizeX = vImarisDataSet.GetSizeX;
aSizeY = vImarisDataSet.GetSizeY;
vSizeT = vImarisDataSet.GetSizeT;
%% %Get the Size of the Time Dimension
vDataSize = [vDataSize, vImarisDataSet.GetSizeT];

%% % Here only the first time point is considered to show the principle
vSizeT = vImarisDataSet.GetSizeT;
for vIndexT = 1:vSizeT-1
%% %extract the Mask Data from Imaris

vMaskDataSet = vImarisObject.GetMask( ...
 vDataMin(1), vDataMin(2), vDataMin(3), ...
 vDataMax(1), vDataMax(2), vDataMax(3), ...
 vDataSize(1), vDataSize(2), vDataSize(3), vIndexT);
end
%% %Convert the Imaris data into an Matlab Matrix
for vIndexT = 1:vSizeT-1
vMask = GetDataVolume(vMaskDataSet, 0, vIndexT);
end

And here is the error I get when I run the modified code:

Error using GetDataVolume (line 23)
Java exception occurred: 
Imaris.Error

 mType = "bpIllegalRangeException"

 mDescription = "Error in bpImageT<Element>::CopyFromChannel"

 mLocation = "bpIceDataSet::GetDataSubVolumeAs1DArrayBytes"

 at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method)

 at sun.reflect.NativeConstructorAccessorImpl.newInstance(Unknown Source)

 at sun.reflect.DelegatingConstructorAccessorImpl.newInstance(Unknown Source)

 at java.lang.reflect.Constructor.newInstance(Unknown Source)

 at java.lang.Class.newInstance(Unknown Source)

 at IceInternal.BasicStream.createUserException(BasicStream.java:2779)

 at IceInternal.BasicStream.access$300(BasicStream.java:14)

 at IceInternal.BasicStream$EncapsDecoder10.throwException(BasicStream.java:3298)

 at IceInternal.BasicStream.throwException(BasicStream.java:2291)

 at IceInternal.OutgoingAsync.throwUserException(OutgoingAsync.java:399)

 at Imaris.IDataSetPrxHelper.end_GetDataSubVolumeAs1DArrayBytes(IDataSetPrxHelper.java:4248)

 at Imaris.IDataSetPrxHelper.GetDataSubVolumeAs1DArrayBytes(IDataSetPrxHelper.java:4078)

 at Imaris.IDataSetPrxHelper.GetDataSubVolumeAs1DArrayBytes(IDataSetPrxHelper.java:4065)


Error in TestMask (line 52)
vMask = GetDataVolume(vMaskDataSet, 0, vIndexT);

If the code runs correctly I expect it will return a time series made of 61 binary images (times frames). From this I would then extract 2D statistics for each surface, in each time frame.

I know you've heard that before but I'm very new to this and my knowledge of Matlab being very limited, I think it is likely the answer is not far off, but I just can't see it.

If you think I am simply going at it the wrong way, and that there may be a much easier solution to my problem, your suggestion would be greatly appreciated too!

Best wishes,

Florian