Objectives
A benefit of the digital environment is that of data transport. At present, if a Breast Care
Unit is short staffed, the movement of data for reading to be performed at another centre would be
difficult. The eDiaMoND infrastructure would enable easy movement of information with high levels
of security to be able to share workloads. It would be practical with this solution to request second
readings from another radiologist from another trust, or seek arbitration services.
Image processing and data mining are aspects of digital mammography which could offer benefits to
clinicians to aid in both the detection of abnormal features as well as in the diagnosis of cancers.
Standardisation of images from different centres in the UK would enable a database to be built using
scans taken on different machines or at different sites, with the effect being that they would all
appear as if they were produced on the same machine under the same conditions. This would enable temporal
comparisons to be performed , which if coupled with methods for removing known changes with age, HRT and
other contributing factors, would indicate abnormal changes over time.
Data mining techniques could also be used to prompt radiologist for likely areas of interest on images
and could possibly be considered for second or arbitration reading.
Computer aided training could enable the sharing of interesting cases across the UK as well as a
removing the need for intensive use of senior radiologists time to train more junior staff. Training
would also be auditable thus enabling the Breast Screening Programme and Breast Imaging services in
general to monitor performance of staff more closely.
With the above vision in mind, the project aims to develop a prototype system which, demonstrates
some of the key features of the future digital world, as well as providing a resource of anonymised
breast imaging data to aid in the development of innovative techniques to assist with detecting and
diagnosing breast cancer.
|