Beehives and Keynotes

Once you get through the sales pitch to the shareholders and wall street analysts we get to the meat of this mornings key note  - the announcement of Oracle Beehive a new updated version of Collaboration Suite.  Oracle demonstrated the product to me while I worked at DFID back in 2004.

All in all it looks a great product, it is based upon open standards it should be possible to link it in to existing infrastructure. I'll try and find out more about the product of the next few days - in particular what is different between Beehive and the old Collaboration suite.

Web 2.0 and Oracle

One of the uncomfortable truths for Oracle is that the Web 2.0 world runs on MYSQL.

There a undoubted problems with this but when Digg, Yahoo  and Google use the LAMP stack to build and deploy large scale performant applications which require no direct licence payment - idea for cash limited startups.

As programs such as Twitter becomes moneytisied by being deployed into companies as collaboration tools, they will come with MYSQL - not Oracle. This could be a major threat as DBA become use to supporting MYSQL and the bottom line costs become significantly cheaper, CIO & CTO become more aware of use of OS solutions.  

DICOM support in Oracle 11g - Case study Novartis.

It takes 15  years to develop new drugs. Companies want to improve the critical path. From the compound selection to its final certification for clinical use.

The use of imaging is becoming more critical to show success of new drugs as they show the effects of the trial drugs quicker than traditional physical diagnostics.

Hospitals become part of the trail and are contracted as 3rd parties. They take modiality scans of patients during the trial and send scans back to the drug company for study management.

The key challenges the companies have when trying to manage the trails evidence using medical images include:

• No single RIS / PACS are hospital centric, no workflow quality management  solution
• Data Access and Control
• No easy access to data for follow up after initial work
• Scattered environment , prop images
• Data transfer
• No single standards, no audit trail, patient confidentially ,
• data

Novartis upgraded to 11g in 2008 

Images are directed as an object in the DB. Its no longer a blob. Data does not need to be parses, so file infrastructure is not needed. The DB manages the archive as part its data life cycle.

Future a requirements is for the support of a flexible  DICOM dictionary - managed through the database metadata service.

The development of a data quality rules through a XML schema. This would include image level rules, content of images, audit. This can then be brought