The degradation of performance of a state-of-the-art skin image classifier when applied to patient-driven internet search

Seung Seog Han, Cristian Navarrete-Dechent, Konstantinos Liopyris, Myoung Shin Kim, Gyeong Hun Park, Sang Seok Woo, Juhyun Park, Jung Won Shin, Bo Ri Kim, Min Jae Kim, Francisca Donoso, Francisco Villanueva, Cristian Ramirez, Sung Eun Chang, Allan Halpern, Seong Hwan Kim, Jung Im Na

Research output: Contribution to journalArticlepeer-review

Abstract

Model Dermatology (https://modelderm.com; Build2021) is a publicly testable neural network that can classify 184 skin disorders. We aimed to investigate whether our algorithm can classify clinical images of an Internet community along with tertiary care center datasets. Consecutive images from an Internet skin cancer community (‘RD’ dataset, 1,282 images posted between 25 January 2020 to 30 July 2021; https://reddit.com/r/melanoma) were analyzed retrospectively, along with hospital datasets (Edinburgh dataset, 1,300 images; SNU dataset, 2,101 images; TeleDerm dataset, 340 consecutive images). The algorithm’s performance was equivalent to that of dermatologists in the curated clinical datasets (Edinburgh and SNU datasets). However, its performance deteriorated in the RD and TeleDerm datasets because of insufficient image quality and the presence of out-of-distribution disorders, respectively. For the RD dataset, the algorithm’s Top-1/3 accuracy (39.2%/67.2%) and AUC (0.800) were equivalent to that of general physicians (36.8%/52.9%). It was more accurate than that of the laypersons using random Internet searches (19.2%/24.4%). The Top-1/3 accuracy was affected by inadequate image quality (adequate = 43.2%/71.3% versus inadequate = 32.9%/60.8%), whereas participant performance did not deteriorate (adequate = 35.8%/52.7% vs. inadequate = 38.4%/53.3%). In this report, the algorithm performance was significantly affected by the change of the intended settings, which implies that AI algorithms at dermatologist-level, in-distribution setting, may not be able to show the same level of performance in with out-of-distribution settings.

Original languageEnglish
Article number16260
JournalScientific Reports
Volume12
Issue number1
DOIs
StatePublished - Dec 2022

Fingerprint

Dive into the research topics of 'The degradation of performance of a state-of-the-art skin image classifier when applied to patient-driven internet search'. Together they form a unique fingerprint.

Cite this