Apply cellpose to RGB images?

I'm trying to finetune cellpose for nuclear segmentation of histology images in MATLAB R2025a. I am able to train the model using RGB tiles and corresponding 2D mask images, but am unable to sample RGB images using segmentCells2D().

Training code:

trainCellpose(pthTraining, ...
    OutFile,...
    PreTrainedModel="nuclei", ...
    MaxEpoch=100, ... 
    ImageSuffix="_im", ...
    LabelSuffix="_mask", ...
    GPUBatchSize=24);

% load the model

cell_size=17;
cp=cellpose(Model=OutFile);

For im = [256x256x3 uint8], this code succeeds:

labels=segmentCells2D(cp,rgb2gray(im),ImageCellDiameter=cell_size);

While this code returns an error:

labels=segmentCells2D(cp,im,ImageCellDiameter=cell_size);

Error using cellpose/segmentCells2D

labels=segmentCells2D(cp,im,ImageCellDiameter=cell_size);

^^

Invalid argument at position 2. Value must be a matrix."

Question: Can I segment RGB images with cellpose in Matlab? The color in histology images is important in distinguishing blue nuclei from non-nuclear round objects, such as shadows or red blood cells.

Answer 1

[Cris LaPierre]

It looks like segementCells2D only supports intensity images.

See the image input description here in the doc.

[Cris LaPierre]

Also, from the Getting Started with Cellpose page: "Cellpose models expect a 2-D intensity image as input. If you have an RGB or multichannel image, you can convert it to a grayscale image by using the im2gray function, or extract a specific color channel by using the imsplit function."