How to use this model for multi-modality dataset?

[yuanyuan29]

Thanks for your excellent work. I am wondering how can I use this model to deal with multi-modality data?

[DISAPPEARED13]

It seems that the input channel in model definition is 1, single modality.
I'm curious if this self-supervised pre-training based weights can identify information about other modalities(for example T1/T2/Flair/... in MRI, Arterial/Venous/Delay/... in CECT)