Can feature extractors meant for 2D images be used with 3D volumes for purpose of classification?

I have to classify and segment some regions in 3D volumes (source is confocal microscopy and MRI - the images are stacked slices in .tif format).

I have experience with classification of images in 2D and have used SIFT, SURF, LBP, LTP, DCT etc to extract features from 2D images. Once the features were extracted I could use a number of classifiers such as SVM, KNN, Kmeans etc to classify the features.

Now the thing is that I dont know how to do the same in 3D. How do these feature extractors translate to 3D volumes? Are there specific 3D feature extractors or should I use the same on every slice and concatenate the features depth wise?

I'm very new to 3D image classification and would appreciate some help or anyone that could point me in the correct direction. As there does not seem to be much help online on the subject. Most of the successfull MRI classification is done using CNN whereas I am looking to classify using conventional machine learning.

Thank you

Answer 1

[Shubh Sahu]

hey!

here are the few links which will be helpful for you

https://www.mathworks.com/help/images/3d-volumetric-image-processing.html

https://www.mathworks.com/help/images/segment-3d-brain-tumor-using-deep-learning.html

https://www.mathworks.com/help/images/segment-lungs-from-3-d-chest-mri-data.html