Perceptual learning in dermatology-A Finnish cohort study of undergraduate medical students

Implicit visual skills play an important role in the recognition of skin-related conditions.

We aimed to evaluate effectiveness and practicability of digital perceptual learning modules (PLMs) during undergraduate teaching of dermatology.

The study consisted of four subsequent dermatology courses including 105 medical students. PLMs designed for an online setting were carried out before, during and at course ends, as well as 6-12 months after the courses (N = 33). We investigated four important outcome measures regarding perceptual learning: diagnostic accuracy (%-percentage of correct responses), decision duration (response time), recognized features (decision basis) and student-perceived confidence.

Diagnostic accuracy (p < 0.001, effect size $$  = 0.82), fluency (p < 0.001, $$  = 0.23) and confidence (p < 0.001, $$  = 0.74) increased significantly with successive PLMs during courses. Students classified more visual features and based the diagnosis more on primary lesion. Accuracy increased in all tasks during the courses and reached over 90% in diagnoses of the first to third task difficulty quartile. In the most difficult quartile accuracy reached to 60%. In the follow-up, students' performance remained at high level. Analysis of diagnostic errors showed that there were specific conditions which were systematically confused with each other.

Digital PLMs improved high rates of diagnostic accuracy, fluency and student-perceived confidence in recognition of skin-related conditions. There was a long-term consistency in high performance suggesting effective learning retention. In the digital setting, PLMs were practicable and easily integrated into traditional teachings. We believe that there is extensive potential for a wider use of perceptual learning to improve nonanalytical visual skills in dermatology and medical education in general.