Why would scientists share data?

F1000Research is an open access journal for life scientists, with rapid publication, transparent peer review (post-publication) and full data deposition and sharing. Open Access Journal have been widely discussed. I have talked about it here. However, sharing data is barely discussed. This journal has recently shared their experience with data openness:
http://blog.f1000research.com/2013/03/14/data-why-openness-and-sharing-are-important/

They raised several important questions, such as verification of results/conclusions, attracting collaborators or establish that you were the first to obtain those data. Scientists have also been questioned about their views and reported in Data Sharing by Scientists: Practices and Perceptions.

As a researcher, I have sometimes data which I would like to share (patient data can't be shared of course...), but I am torn between "I can't write a paper about the data itself" and "If I share it without caution, someone will use my data and not even acknowledge me and it gave me so much work!".

Did you ever considered sharing your data?

Databases of Medical Images

Medical Image Databases are great to work on medical imaging processing and sometimes to get a feeling of the images before using real data. I have found some databases in the internet that I consider quite interesting:


Minimal Interval Resonance Imaging in Alzheimer's Disease MIRIAD
This database is focused on Alzheimer Diseases patients, but also control patients.

Mini Mammographic Image Analysis Society Database Mini-MIAS
A database with mammograms and with medical information.

Medpix Teaching database MedPix
A database with examples to teach from CT to MRI.

Retrospective Image registration Evaluation RIRE
A database with images to register and the ground truth to compare.

Alzheimer's Disease Neuroimaging Initiative ADNI
This a huge database of patients images with PET and MRI.

Do you know other good databases?