The implementation for converting 2D slices to 3D above looks correct. However, you could change the initialization of “image3d” to:
image3d = zeros(rows, columns, 100, 'uint8'); % Initialize 3D array with correct size
Refer to this link for different methods such as using "cat" to achieve the same functionality : https://www.mathworks.com/matlabcentral/answers/154714-how-can-i-convert-2d-images-to-a-3d-image
To visualize and process Regions of Interest (ROI) and skin layer regions in images using MATLAB refer to this link:
To perform automatic detection of volumes of specific types, you can adopt deep learning methods like "unet-3d," which is a "CNN" architecture designed for image segmentation with a symmetric encoder-decoder structure and skip connections, more information here :
