Learning-by-Concordance of Perception: A Novel way to Learn to Read Thoracic Images

Abstract

Learning to interpret thoracic images requires intensive instructor support. Given current cohort sizes at teaching hospitals in North America, instructor availability is rare. A Learning-by-concordance of perception (LbCP) online tool was introduced in a second-year course on lung and oxygenation. The LbCP tool presents thoracic images, students must point or outline abnormal structures directly on the screen and name the lesion. Thereafter, images with correct outline are superimposed on student's work and three key-messages are provided. We aimed to measure student perception of LbCP tool's usefulness and ease of use. The online tool was developed and implemented for second year students for cohorts in 2016, 2017 and 2018 (n=296; 303; and 280; N=879). A survey, comprisingsix questions on a Likert scale was designed to measure perceptions about tool utility and ease of use. An ANOVA analysis was carried out to ensure the normality of the data, and a principal axis factor analysis was used to confirm the presence of the two expected clusters corresponding to our two dimensions. The ANOVA conducted on the combined three year data set revealed an F value of 7.688 (p=0.001), and principal axis factorial analysis revealed a one factor solution. The percentage of variance explained by the factor was 44.5%, with factor loadings leaning heavily in favor of the tool's perceived utility. A second factor was just shy of the eigenvalue threshold of 1.0 and could provide support for the tool's ease of use. The online LbCP tool shows promising impact over three cohorts of students in three consecutive years. Students recognize the pedagogical value of the tool and express their willingness to use more of it in their training.