SUPPLEMENTAL DATA CONTENT DESCRIPTION s1.csv Description: Characteristics of patients enrolled in the study. Data elements: patient_id ISIC Archive patient identifier fst Patient's Fitzpatrick Skin Type based on in-clinic questionnaire race Patient's self-reported race ethnicity Patient's self-reported ethnicity eye_color Patient's eye color hair Patient's hair color dermatoscope Type of dermatoscopic device used during the visit patient_sex Patient's biologic sex patient_age_approx Patient's age in 5-year bins s2.csv Description: Characteristics of skin sites analyzed in the study. Data elements: tag_id Study local "tag" identifier for each unique skin site patient_id ISIC Archive patient identifier lesion_id ISIC Archive lesion identifier type Distinguishes between lesion or normal-skin anatomic_site Location on the patient's body s3.csv Description: Analysis of inter-rater agreement for in-person Pantone and MST on both lesional and non-lesional sites. Data elements: tag_id Study local "tag" identifier for each unique skin site type Distinguishes between lesion or normal-skin anatomic_site Location on the patient's body mst_r1 Rater 1's in-person MST rating mst_r2 Rater 2's in-person MST rating pantone_r1 Rater 1's in-person Pantone rating pantone_r2 Rater 2's in-person Pantone rating pantone_r1_Red_Yellow Rater 1's in-person Pantone rating; Red-Yellow component pantone_r2_Red_Yellow Rater 2's in-person Pantone rating; Red-Yellow component pantone_r1_Dark_Light Rater 1's in-person Pantone rating; dark-light component pantone_r2_Dark_Light Rater 2's in-person Pantone rating; dark-light component s4.csv Description: Analysis of inter-rater agreement for in-person colorimeter device measurements on non-lesional sites. Data elements: tag_id Study local "tag" identifier for each unique skin site type Distinguishes between lesion or normal-skin anatomic_site Location on the patient's body l_1 "L*"-channel of 1st colorimeter assessment l_2 "L*"-channel of 2nd colorimeter assessment l_3 "L*"-channel of 3rd colorimeter assessment average_l Average "L*"-channel across all assessments a_1 "a*"-channel of 1st colorimeter assessment a_2 "a*"-channel of 2nd colorimeter assessment a_3 "a*"-channel of 3rd colorimeter assessment average_ita Average "a*"-channel across all assessments b_1 "b*"-channel of 1st colorimeter assessment b_2 "b*"-channel of 2nd colorimeter assessment b_3 "b*"-channel of 3rd colorimeter assessment average_b Average "b*"-channel across all assessments ita_1 ITA angle of 1st colorimeter assessment ita_2 ITA angle of 2nd colorimeter assessment ita_3 ITA angle of 3rd colorimeter assessment average_ita Average ITA angle across all assessments s5.csv Description: Analysis of clustering of FST, MST, and Pantone classes in the 2-dimensional CIELAB color space on non-lesional sites. Data elements: tag_id Study local "tag" identifier for each unique skin site fst Patient's Fitzpatrick Skin Type based on in-clinic questionnaire average_l Average "L*"-channel across all colorimeter assessments average_a Average "a*"-channel across all colorimeter assessments average_b Average "b*"-channel across all colorimeter assessments mst_r1 Rater 1's in-person MST rating mst_r2 Rater 2's in-person MST rating pantone_r1 Rater 1's in-person Pantone rating pantone_r2 Rater 2's in-person Pantone rating s6.csv Description: Analysis of concordance between MST ratings of non-lesional sites performed by Rater 1 in-person versus thorugh 3D total body photographs (TBP). Data elements: tag_id Study local "tag" identifier for each unique skin site type Distinguishes between lesion or normal-skin anatomic_site Location on the patient's body tbp_lighting Lighting setting of TBP capture (white "w" or cross-polarized "xp") mst_r1_inoffice Rater 1's in-person MST rating mst_tbp Rater 1's TBP-based MST rating s7.csv Description: Analysis of colorimeter vs. image-extracted ITA in non-lesional sites. Data elements: isic_id ISIC Archive image identifer dermoscopic_type Dermoscopic type of the image (polarized/non-polarized contact/non-contact) img_l Median "L*"-channel of the image img_a Median "a*"-channel of the image img_b Median "b*"-channel of the image img_ita Image ITA derived from img_l and img_b through the formula ITA = tan-1((L*-50)/b*) * (180/pi) tag_id Study local "tag" identifier for each unique skin site type Distinguishes between lesion or normal-skin anatomic_site Location on the patient's body l_1 "L*"-channel of 1st colorimeter assessment l_2 "L*"-channel of 2nd colorimeter assessment l_3 "L*"-channel of 3rd colorimeter assessment average_l Average "L*"-channel across all assessments a_1 "a*"-channel of 1st colorimeter assessment a_2 "a*"-channel of 2nd colorimeter assessment a_3 "a*"-channel of 3rd colorimeter assessment average_a Average "a*"-channel across all assessments b_1 "b*"-channel of 1st colorimeter assessment b_2 "b*"-channel of 2nd colorimeter assessment b_3 "b*"-channel of 3rd colorimeter assessment average_b Average "b*"-channel across all assessments ita_1 ITA angle of 1st colorimeter assessment ita_2 ITA angle of 2nd colorimeter assessment ita_3 ITA angle of 3rd colorimeter assessment average_ita Average ITA angle across all assessments s8.csv Description: Analysis of crowd-sourced FST vs. in-person FST rating. Data elements: isic_id ISIC Archive image identifer tag_id Study local "tag" identifier for each unique skin site fst Patient's Fitzpatrick Skin Type based on in-clinic questionnaire type Distinguishes between lesion or normal-skin anatomic_site Location on the patient's body dermoscopic_type Dermoscopic type of the image (polarized/non-polarized contact/non-contact) crowd_has_consensus Binary indication of whether enough reads were collected to attribute a label crowd_qualified_reads Frequency of crowd reads for a given image crowd_majority_label Crowd consensus FST label (majority label is defined as the FST label selected by at least 50% of qualified reads or the label selected by the majority if more than 12 qualified reads were collected) s9.csv Description: Analysis of estimated risk scores from an AI algorithm (ADAE) Data elements: isic_id ISIC Archive image identifer dermoscopic_type Dermoscopic type of the image (polarized/non-polarized contact/non-contact) tag_id Study local "tag" identifier for each unique skin site mst_r1 Rater 1's in-person MST rating mst_r2 Rater 2's in-person MST rating fst Patient's Fitzpatrick Skin Type based on in-clinic questionnaire adae AI-assessed malignancy risk score