Grant Number:	1R37CA262110-01A1 Interpret this number
Primary Investigator:	Salkowski, Lonie
Organization:	University Of Wisconsin-Madison
Project Title:	Defining and Optimizing Critical Interpretation Skills in Screening Mammography to Improve Cancer Detection
Fiscal Year:	2022

Project Summary Abstract There is a critical lack of information regarding how radiology residents develop interpretive skills for screening mammography. Consequently, little is known how to avert or correct errors in perception and cognition. The majority of radiology residents have only the required minimum 12 weeks of breast imaging training. Yet, the majority of mammograms in the United States are interpreted by general radiologists, not board-certified breast imagers. The negative impact on patient care by not filling this gap in education is continued missed cancers and unnecessary additional imaging and biopsies. The long-term goal of this project is to reduce interpretive errors in breast imaging for improved patient outcomes by providing objective tools and resources to improve trainee education. The overall objectives of this proposal are to understand cognitive and perceptual skill development in a specialized simulation system by bringing together experts in radiology, computer science, educational psychology, and psychology. In residency there is no opportunity for residents to interpret mammographic studies independently from attending radiologists. Thus, offering no opportunity to ‘practice’ their skills. Our education system needs improvement. This project will implement a specialized simulation system with sequential evaluation of radiology resident performance in screening mammography providing objective measures and feedback to trainees. The central hypothesis of this project is that a systemized simulation screening mammography system within residency training will aid in identifying critical cognitive and perceptual errors that negatively impact patient care and outcomes. This hypothesis will be tested by three specific aims: 1) Use machine learning tools to develop congruent case sets for consecutive simulation sessions. 2) Assess the impact of sequential simulation training and examine predictors of individual differences in learning trajectories, by a) assess interpretive skills of radiology residents and fellows in the breast screening simulation system, b) assess real-time perception skills with eye tracking in simulation sessions, and c) assess the resident’s fundamental cognitive and perceptual abilities as well as personality traits. 3) Assess and quantify the effects of sequential simulation training. This research proposal is innovative, in that it provides sequential opportunities for radiology residents to independently interpret screening mammograms in a specialized simulation system that will obtain information about the development of interpretive and perceptive skills over time during real-world practice. This research proposal is significant because simulation training is effective at taking high-risk stakes skills and putting them in a low risk encounters to assess and improve performance. Improving the preparation of radiology residents to interpret screening mammograms will have far reaching impact on women’s health and overall public health.





None