The Royal College of Radiologists National Imaging Archive Concept
An ambitious project carried out in collaboration with The Royal College of Radiologists (UK) to allow radiologists to send DICOM data remotely from any PACS to a centrally hosted National Image Data Archive.
data • Benefit to RCR members as a national method of storing teaching case data • Intellectual property goldmine ◦ AI / CAD ◦ Sale to non-members for study • The defacto archive for special use cases ◦ FRCR exam cases ◦ READ safety cases ◦ Specialist interest cases • New standard universal method to submit cases to RCR Concepts to consider
& Harvey at Informatics Committee meeting • Challenge set by Nicola Strickland ◦ “Why can’t you get on and make it then?” ◦ Shot ourselves in the feet? • Discussion in The Ship Tavern (very close to RCR) ◦ Healthflow (DF) - frontend development ◦ Medical Connections (DH) - DICOM / backend development ◦ Both companies offered team expertise and development time ◦ Collaborative rapid wireframe development for meeting with vendors RCR National Archive: How we got talked into this!
Server End user application Right click “Send to National Archive Modal Dialog popover Fields for TCE data set input by author Local Anonymisation Export Manager
Server End user application Right click “Send to National Archive Modal Dialog popover Fields for TCE data set input by author Local Anonymisation Export Manager RCR TCE Listener Anonymous Image Data Store TCE case data store
Server End user application Right click “Send to National Archive Modal Dialog popover Fields for TCE data set input by author Local Anonymisation Export Manager RCR TCE Listener Anonymous Image Data Store TCE case data store Case data editable via web application by author
automatically purged Completed case • Data committed to archive • Can be made public by author • Default kept in private collection 30 days to complete
by JSON data • Application behaviour not like a website • Instant response for ◦ Searching ◦ Filtering ◦ Loading individual case data • Technology Stack ◦ Node.js ◦ Angular.js ◦ Healthflow Layout Engine
experience ◦ No waiting around for search / load / filter actions ◦ Works on anything with a modern browser • Sensible feature set ◦ Powerful realtime full text search ◦ Realtime table column filtering ◦ ‘My Cases’ ◦ Multiple organisation methods ▪ Starred / favourite cases ▪ Custom lists (can be private or public) • Permissions system ◦ Case and list privacy