To the Editor:
The Fitzpatrick scale (FST) was developed to classify skin types based on responses to ultraviolet light, but is now commonly used in dermatology as a surrogate for skin tone. The FST does not adequately account for the diversity in patients with skin of color (SOC), however, limiting its clinical utility.1,2,3 In response to these shortcomings, the Monk Skin Tone (MST) Scale, a 10-shade grading scale, was developed as a more inclusive skin tone assessment.4 However, the relationship between FST and MST is not well known. While clinical validation of MST is limited, there are efforts by Google to use this scale to help train their artificial intelligence (AI) models. Recently, a partnership between Google and Stanford researchers generated the Skin Condition Image Network (SCIN) dataset, which includes extensive dermatological images from patients recruited from Google search advertisements.5 Our study utilizes this dataset to benchmark the relationship between FST and MST across different evaluators, ethnic backgrounds, and skin types.
The Fitzpatrick scale (FST) was developed to classify skin types based on responses to ultraviolet light, but is now commonly used in dermatology as a surrogate for skin tone. The FST does not adequately account for the diversity in patients with skin of color (SOC), however, limiting its clinical utility.1,2,3 In response to these shortcomings, the Monk Skin Tone (MST) Scale, a 10-shade grading scale, was developed as a more inclusive skin tone assessment.4 However, the relationship between FST and MST is not well known. While clinical validation of MST is limited, there are efforts by Google to use this scale to help train their artificial intelligence (AI) models. Recently, a partnership between Google and Stanford researchers generated the Skin Condition Image Network (SCIN) dataset, which includes extensive dermatological images from patients recruited from Google search advertisements.5 Our study utilizes this dataset to benchmark the relationship between FST and MST across different evaluators, ethnic backgrounds, and skin types.
