Dianne Georgian-Smith MD, Presenter: Consultant, Vizovo Researcher, Analogics, Inc Researcher, Siemens AG

John Ryan PhD, Abstract Co-Author: Director, Vizovo Ltd and Vizovo Inc

Eren D. Yeh MD, Abstract Co-Author: Nothing to Disclose

Lorraine B. Smith MBBCh, Abstract Co-Author: Nothing to Disclose

Julie Ann Ritner MD, Abstract Co-Author: Nothing to Disclose

Sara Mannarelli Durfee MD, Abstract Co-Author: Nothing to Disclose

Hope Elizabeth Peters MD, Abstract Co-Author: Nothing to Disclose

Robyn L. Birdwell MD, Abstract Co-Author: Nothing to Disclose

The goal was to determine criteria leading to the perception of architectural distortion (AD) from which a teaching module was developed. The effectiveness of this module was tested on trainees.

Determining Criteria of AD: A Test Set of 50 cases (198 images) was assembled with both positive AD/ negative cases; some had more than one site of AD. Five experienced radiologists completed the Test Set, and each was independently asked “what they saw” on selected cases. From these criteria a tutorial was developed. Self- Tutorial: A powerpoint tutorial illustrated criteria for AD and methods to distinguish AD from overlapping tissues; a second bank of 20 images was produced for practice and provided immediate feedback and the option of reviewing one’s calls vs. truth. Testing effectiveness of tutorial: 4 trainees completed the three phases of this study: the Test Set (50 cases), then the tutorial, and repeat of the Test Set.

1. TEST SET- NO FEEDBACK: 5 experts and 15 trainees completed the Test Set once. The experts performed better than the trainees. The highest ROC 0.96 (expert) and lowest 0.473 (trainee); 2. PRACTICE SET WITH FEEDBACK: Of 4 trainees who completed the Practice set and repeat Test Set, 2/4 reviewed practice cases more than once; both improved ROC's;3. REPEAT TEST SET AFTER TUTORIAL: ROC and sensitivities decreased 3/4 trainees; specificity increased in 4/4 trainees

1. Experienced radiologists perform better than trainees in seeing AD. 2. In preliminary data, the largest impact of the tutorial is the improvement in specificity, but at the expense of decreased sensitivity, as has been noted in screening studies. 3. The data may be skewed since this is a small sample size; further testing is planned. Study was supported by RSNA Educational Seed Grant.

This study illustrates the importance of testing modules and the importance of understanding the perception of architectural distortion.

Georgian-Smith, D, Ryan, J, Yeh, E, Smith, L, Ritner, J, Durfee, S, Peters, H, Birdwell, R, Mammographic Architectural Distortion: Understanding the Perception to Develop Methods for Teaching.  Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL. http://archive.rsna.org/2011/11014086.html