Repeated errors trying to install synthseg

I have been trying to install and run synthseg on Matlab R2022a without success. I've had a look at the questions on this community and have tried a couple of fixes but it does not seem to be working despite these. I have installed the add on for deep learning converter to tensorflow.

This is the error code I'm getting:

zipFile = matlab.internal.examples.downloadSupportFile("image","data/brainSegData.zip");

filepath = fileparts(zipFile);

unzip(zipFile,filepath)

>> dataDir = fullfile(filepath,"brainSegData");

>> trainedBrainCANDINetwork_url = "https://www.mathworks.com/supportfiles/"+ ...

"image/data/trainedSynthSegModel.zip";

downloadTrainedNetwork(trainedBrainCANDINetwork_url,dataDir)

>> net = importNetworkFromTensorFlow(fullfile(dataDir,"trainedSynthSegModel"))

Unrecognized function or variable 'importNetworkFromTensorFlow'.

I then tried using this older expression because I have an old OS and older version of matlab:

>> net = importTensorFlowNetwork(fullfile(dataDir,"trainedSynthSegModel"))

Error using nnet.internal.cnn.tensorflow.importTensorFlowNetwork

ModelFolder not found:

'/Users/clarabelessiotis/Documents/MATLAB/Examples/R2022a/supportfiles/image/data/brainSegData/trainedSynthSegModel'.

Error in importTensorFlowNetwork (line 107)

Network = nnet.internal.cnn.tensorflow.importTensorFlowNetwork(modelFolder, varargin{:});

>>

Answer 1

[Abhinav Aravindan]

Hi Clara,

I assume that you are using the “Brain MRI Segmentation Using Pretrained 3-D U-Net Network” example from the “Medical Imaging Toolbox”. The “Medical Imaging Toolbox” appears to have been introduced in R2022b as per the release notes below:

https://www.mathworks.com/help/medical-imaging/release-notes.html#:~:text=Introducing%20Medical%20Imaging%20Toolbox%3A%20Import%2C%20visualize%2C%20register%2C%20label%2C%20and%20segment%202%2DD%20and%203%2DD%20medical%20images

To use the example, you may consider upgrading to MATLAB R2022b or later and access the example using the following command:

openExample('images_deeplearning/BrainMRISegmentationUsingTrained3DUNetExample') 

For further details on the example, please refer to the documentation below from the MATLAB R2022b release. Please note that this example uses the following Toolboxes in addition to the “Medical Imaging Toolbox”:

1. Deep Learning Toolbox Converter for TensorFlow Models
2. Deep Learning Toolbox
3. Image Processing Toolbox
4. Computer Vision Toolbox

https://www.mathworks.com/help/releases/R2022b/medical-imaging/ug/Brain-MRI-Segmentation-Using-Trained-3-D-U-Net.html

I hope this answers your query!