Abstract
Although a deadly form of skin cancer, melanoma is treatable if detected early. Existing approaches in melanoma detection training employ a rule-based method where lesions are assessed by their Asymmetry, Border, Color, Diameter and Evolvement in appearance (i.e., the ABCDE rule). However, the rule-based training practices in melanoma detection were not effective. In the current study, we assessed an innovative way to train melanoma detection using the principles of perceptual expertise training. All participants first reviewed the ABCDE rules pamphlet, and were then given the Melanoma Detection Test (MDT) as the pre-test where they categorized a set of skin lesion images as either "melanoma" or "benign". Participants in the perceptual expertise training group received four training sessions where they were taught to categorize melanoma and benign lesions to 95% accuracy. Participants in the control group received no training. After training, all participants were retested with the same items on the MDT. As compared to the control group, the training group showed significant improvement in melanoma detection and a shifted response criterion from liberal (i.e., bias toward categorizing a lesion as melanoma) to neutral, and both improvements maintained a week after the training. These findings indicate that perceptual expertise training is a promising approach to train melanoma detection.
Meeting abstract presented at VSS 2016