Yan Chen, Presenter: Nothing to Disclose

Hazel Janet Scott, Abstract Co-Author: Nothing to Disclose

Jonathan James BMBS, Abstract Co-Author: Nothing to Disclose

Andrew Evans MRCP, FRCR, Abstract Co-Author: Nothing to Disclose

Alastair G. Gale PhD, Abstract Co-Author: Nothing to Disclose

Determining whether mammographic experience and performance are related to particular visual search behaviour so as to develop new training programmes.

12 mammographers were divided into two groups, according to screening experience. For each group (n=6), measures of specificity and small cancer detection performances were obtained for both their real-life screening and performance on a national UK self-assessment program. Subsequently, they examined 40 two view screening cases on a digital workstation whilst their visual search behaviour was recorded. For each case, participants rated their confidence in whether an abnormality was present; and if so specified its location and mammographic feature type. Individuals’ visual search data were analysed and where errors were made these were classified as due to search, detection or interpretation factors. Qualitative and quantitative differences between the groups were compared.

A one-way ANOVA showed the more experienced group was significantly better at cancer detection on the self-assessment task [F(1, 23)=5.4,p<.05] and descriptively better for other cancer detection measures (p=n.s.). Importantly, differences in visual search behaviour between the groups were found with experienced observers making fewer eye movements, fixating faster but dwelling longer on key abnormalities whereas less experienced observers made more eye movements with significantly shorter fixation duration.

A relationship was found between experience and visual search behaviour which demonstrates the potential of utilising visual search data during mammographic interpretation training.

The increasing use of digital mammography supports the potential use of eye movement information for future training in interpreting mammograms.

Chen, Y, Scott, H, James, J, Evans, A, Gale, A, Breast Screening: Visual Search as an Aid for Digital Mammographic Interpretation Training.  Radiological Society of North America 2009 Scientific Assembly and Annual Meeting, November 29 - December 4, 2009 ,Chicago IL. http://archive.rsna.org/2009/8004634.html