Frequently Asked Questions
Q. Why should I use NIDB instead of a different database? A. This is the most important question! The short answer is because NIDB is simple and powerful, and easy to install and use. You can start storing, searching and sharing data in less than a day. And it’s free. The long answer is a bit longer. NIDB has been in development since 2005, a time when many labs had collected enough fMRI data to say “So much data! I need my MR data, and how can I search for it and download it?”. This evolved into “Where’s my X modality data, and why isn’t it associated with the MR data?”. Eventually, it evolved into “Alright, I can instantly search for any modality from any of my 10,000 subjects. How about searching for analysis results?” What resulted was an extremely simple yet powerful database that stores any modality data, can instantly retrieve any data, and automatically analyze it. In the end, the long answer comes down to time and money. Whatever size lab, center, or institution you have, you can quickly create a scalable neuroinformatics data store without needing to hire a programmer. You house the data at your site, under your control, and its free.
Q. How big can the database get? A. The Olin Center hosts 18,000 imaging studies from 9,000 subjects. The size of the database is largely dependent on your server hardware. The only practical limit is the unique ID, which is 4 digits followed by 3 letters, for a total of 175 million subjects. If you use 4 digit/4 character, that grows to 4 billion subjects. Beyond that, its up to the hardware and underlying OS to see how big you can grow.
Q. How fast are the queries? A. Queries are generally fast, most returning in less than 0.1sec. Queries returning several thousand series may be on the order of seconds. Speed also depends on the complexity of your query. If you have a lot of criteria, it may go a little slower than if you have less criteria.
Q. Is NIDB secure? A. It depends on your meaning of secure. Should you put the website on a public facing server? Probably not, especially if you plan to store PHI on it. But it is secure enough to be hosted on an intranet. The level of security is also largely dependent on the operating system. Tightening security settings on the server will make the system more secure.
Q. What’s with the UID format? A. We call it the license plate style IDs, because several US states’ plates are in that format. It allows for more UIDs, and slightly easier to transcribe than UIDs of all digits or alternating digits and letters.
Q. Can I help develop NIDB? Or… I have comments on what I’d like to see. Or… I have issues. A. Absolutely! Contact Greg at firstname.lastname@example.org