Open AccessDevelopment of a Checklist Tool to Assess the Quality of Skin Lesion Images Acquired by Consumers Using Sequential Mobile Teledermoscopy
Author(s) -
Uyen Koh,
Brigid BetzStablein,
Montana O’Hara,
Caitlin Horsham,
Clara CurielLewandrowski,
H. Peter Soyer,
Monika Janda
Publication year - 2021
Publication title -
dermatology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.224
H-Index - 92
eISSN - 1421-9832
pISSN - 1018-8665
DOI - 10.1159/000515158
Subject(s) - checklist , kappa , reliability (semiconductor) , medicine , image quality , medical physics , cohen's kappa , skin lesion , quality (philosophy) , inter rater reliability , nuclear medicine , computer science , artificial intelligence , dermatology , statistics , psychology , mathematics , image (mathematics) , rating scale , machine learning , power (physics) , philosophy , physics , geometry , epistemology , quantum mechanics , cognitive psychology
Background: Mobile teledermoscopy is an emerging technology that involves imaging and digitally sending dermoscopic images of skin lesions to a clinician for assessment. High-quality, consistent images are required for accurate telediagnoses when monitoring lesions over time. To date there are no tools to assess the quality of sequential images taken by consumers using mobile teledermoscopy. The purpose of this study was to develop a tool to assess the quality of images acquired by consumers. Methods: Participants imaged skin lesions that they felt were concerning at baseline, 1-, and 2-months. A checklist to assess the quality of consumer sequential imaging of skin lesions was developed based on the International Skin Imaging Collaboration guidelines. A scale was implemented to grade the quality of the images: 0 (low) to 18 (very high). Intra- and inter-reliability of the checklist was assessed using Bland-Altman analysis. Using this checklist, the consistency with which 85 sets of images were scored by 2 evaluators were compared using Kappa statistics. Items with a low Kappa value <0.4 were removed. Results: After reliability testing, 5 of the items were removed due to low Kappa values (<0.4) and the final checklist included 13 items surveying: lesion selection; image orientation; lighting; field of view; focus and depth of view. Participants had a mean age of 41 years (range 19–73), and 67% were female. Most participants (84%, n = 71/85) were able to select and image the correct lesion over time for both the dermoscopic and overview images. Younger participants (<40 years old) scored significantly higher (8.1 ± 2.1) on the imaging checklist compared to older participants (7.1 ± 2.4; p = 0.037). Participants had most difficulty with consistent image orientation. Conclusions: This checklist could be used as a triage tool to filter images acquired by consumers prior to telediagnosis evaluation, which would improve the efficiency and accuracy of teledermatology and teledermoscopy processes. It may also be used to provide feedback to the consumers to improve image acquisition over time.