I attended the first meeting of the UK Osirix User Group which was organised by Dr Mary Roddie of Imperial College, London today. It was pretty well attended for such a seemingly niche topic. The adoption of OsiriX as the chosen method to run the FRCR examination has enforced a standard amongst the progressive radiology educators in our country.
You could broadly split the group into 2 main categories:
- Interested in using OsiriX
- Already using OsiriX (with a wide range of ability and experience in this group)
It was interesting to see some of the notable trainers from around the country already running courses on OsiriX based setups such as Simon Morley (UCL, London) and John Curtis (Aintree, Mersey). But there was also a broad range of interest from College members keen to establish data integrity standards and also tutors from every corner of the UK itching to set up their own digital teaching networks – with a little helping hand.
What we discussed
Undoubtedly the main reason for attending for many today was to see how others are using OsiriX to organise and deliver their teaching.
John Curtis demonstrated how he uses a dual screen setup with an Apple Cinema Display and a Macbook to serve one on one teaching. He can see his database and the answers, whilst the candidate has a large image display of his choosing as well as a minimal toolset to read the image.
Simon Morley showed us remote secure networking methods which allow distance sharing of datasets for teaching purposes. He also agrees with me that a UK wide DICOM teaching server is an attractive and achievable goal. He has already set up a small prototype server which he used to demonstrate retrieval of a Chest radiograph across the 3G network.
Sameer Shamshuddin from Glasgow showed us more rigorous methods of anonymising our DICOM data which go far beyond the stripping abilities of OsiriX. He also demonstrated methods of embedding other files (such as PDF and PPT) within the DICOM container, such that they remain associated to the relevant imaging.
- Data security – we could not agree on whether the patient ID should be retained within the DICOM data and the file encrypted, or the ID data should be totally stripped and a separate database of patient IDs kept securely in the hospital for correlation if needed.
- Case Nomenclature – We agreed that the format for naming cases should follow that already adopted by the RCRXXX001
XXX = 3 letter initials of the contributing radiologist
001 = a case number padded with zeros. The number of zeros is not agreed, RCR currently use just 3bit numbers (999 max cases) which some felt to be too few.
- DICOM fields which we use for labelling cases need to be agreed. Comments is the only OsiriX field which maps to an official DICOM header (StudyComments).
- Tutorials – there was a strong feeling that we need to offer tutorials to engage other UK radiologists in the new process and teach them best practices. Regional courses were an idea to kickstart this. Video tutorials online may be another method, as already demonstrated by some.
The day was a great success overall and showed a level of enthusiasm sufficient to work towards the dream goal of a UK DICOM teaching database.
If there is one take home message it is: Start labelling your studies using the RCR convention and start doing it now. Only use Comments to store your additional diagnostic information.