Let me make sure I understand the question. You have DICOM images from a Siemens scanner that is saved as mosaics:

  https://nipy.org/nibabel/dicom/dicom_mos...

However, dcm2nii is failing to detect them. There are two issues:

 1. Data comes from a Siemens V* MRI software (e.g. VE11). In this case, the mosaic information is stored in the CSA header, and failure to detect this suggests the CSA headers have been stripped from your DICOM. The solution is to search the provenance of your DICOMs and convert images prior to the removal of these tags (which also provide critical sequence infromation). I suspect an anonymization stage removed these details.

   https://nipy.org/nibabel/dicom/siemens_csa.html

 2. Data was acquired using Siemens XA MRI software (e.g. XA30). In this case, the mosaic images are for visual inspection only, and do not contain the information required to analyze the data (e.g. slice timing, gradient information, etc). In this case, you will want to re-export the data from your scanner without the mosaic mode. In my experience, these mosaics lack Type 1 (required) information, so they are not actually valid DICOM images, rather they are DICOM data files. They are too impoverished for reproducible science.

  https://github.com/rordenlab/dcm2niix/is...

The Siemens Research Collaboration Manager affiliated with your center can aid you in these steps.