Skip to content

Improving skin tone determination: a comparison of two skin tone scales

Towards fairness in skin tone determination in AI-algorithms: a comparison of the Fitzpatrick and Monk skin tone scales using spectrophotometry - Towards fairness in skin tone determination in AI-algorithms: a comparison of the Fitzpatrick and Monk skin tone scales using spectrophotometry.

Status
Active, not recruiting
Phases
Unknown
Study type
Observational
Source
NL-OMON
Registry ID
NL-OMON57955
Enrollment
500
Registered
2025-08-19
Start date
2025-08-21
Completion date
Unknown
Last updated
2026-06-15

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

Skin tone

Interventions

Not applicable.

Sponsors

Erasmus MC, Universitair Medisch Centrum Rotterdam
Lead Sponsor

Eligibility

Age
18 Years to 99 Years

Inclusion criteria

Inclusion criteria: In order to be eligible to participate in this study, a subject must meet all of the following criteria:Individuals aged 18 years or older.Willingness to participate and provide written informed consent.Fluent in the Dutch or English language

Exclusion criteria

Exclusion criteria: A potential subject who meets any of the following criteria will be excluded from participation in this study: Individuals unable or unwilling to provide informed consent.Presence of any skin condition that may interfere with spectrophotometry measurements or lead to participant identification from a close-up photograph of the upper and lower arm (e.g., tattoos, burns, or skin disorders in the measurement area).

Design outcomes

Primary

MeasureTime frame
The determination and comparison of the Spearman correlation between an objective measure of skin pigmentation determined by spectrophotometry (individual typology angle (ITA)) and the categories of each skin tone scale (FST, MST).2. The collection of macroscopic and dermatoscopic photos of two unidentifiable sites (forearm, upper arm) for the training and validation of AI algorithms that automatically determine skin tone, phototype and photodamage.

Secondary

MeasureTime frame
3: Determination of differences in means of ITA distributions between adjacent FST and MST categories. An ideal scale would show statistically significant differences between adjacent categories. 4: Quantification of possible gaps in ITA distributions between adjacent FST and MST categories.5: Quantification of uniformity of classification intervals in the FST and MST scales. Performed by calculating the coefficient of variation for the differences in ITAs between adjacent categories for each scale.6: Determination of correlation between questions relevant to photosensitivity (tanning, burning) and the categories of each scale (FST/MST).7: Exploration of the above objectives using participant-reported FST and MST scale categories, as well as by using MST swatches for MST determination.O8: Exploration of the above objectives using the MST groups proposed by regulators: MST 1-4, MST 5-7, MST 8-10. 

Countries

Netherlands

Contacts

Public ContactC Diamantidis

Erasmus MC, Universitair Medisch Centrum Rotterdam

fair-ai@erasmusmc.nl0619751657

Outcome results

None listed

Source: NL-OMON (via WHO ICTRP) · Data processed: Jun 21, 2026