Equity in skin typing: why it's time to replace the Fitzpatrick scale
TL;DR: The Fitzpatrick skin type (FST) scale, with both disproportionate focus on white skin tones and inconsistent use, perpetuates skin color bias as mentioned in this paper, since FST does not purely objectively estimate skin pigmentation, it may inaccurately assess patients regarding risks for skin cancer and from interventions.
Abstract: Dermatologists of color have long championed skin of color representation in education and workforce diversity. For health equity, we must reconsider even fundamental and accepted terminology. The Fitzpatrick skin type (FST) scale, with both disproportionate focus on white skin tones and inconsistent use, perpetuates skin color bias. Indeed, since FST does not purely objectively estimate skin pigmentation, it may inaccurately assess patients regarding risks for skin cancer and from interventions. Dermatology must seek an objective classification system, and given the rise of artificial intelligence (AI), technology-based approaches may be solutions.
Citations
More filters
TL;DR: A consensus statement of 19 members of the International Skin Imaging Collaboration AI working group volunteered to provide a consensus statement to consolidate limited existing literature with expert opinion to guide developers and reviewers of dermatology AI as mentioned in this paper .
Abstract: The use of artificial intelligence (AI) is accelerating in all aspects of medicine and has the potential to transform clinical care and dermatology workflows. However, to develop image-based algorithms for dermatology applications, comprehensive criteria establishing development and performance evaluation standards are required to ensure product fairness, reliability, and safety.To consolidate limited existing literature with expert opinion to guide developers and reviewers of dermatology AI.In this consensus statement, the 19 members of the International Skin Imaging Collaboration AI working group volunteered to provide a consensus statement. A systematic PubMed search was performed of English-language articles published between December 1, 2008, and August 24, 2021, for "artificial intelligence" and "reporting guidelines," as well as other pertinent studies identified by the expert panel. Factors that were viewed as critical to AI development and performance evaluation were included and underwent 2 rounds of electronic discussion to achieve consensus.A checklist of items was developed that outlines best practices of image-based AI development and assessment in dermatology.Clinically effective AI needs to be fair, reliable, and safe; this checklist of best practices will help both developers and reviewers achieve this goal.
35 citations
TL;DR: The Diverse Dermatology Images (DDI) dataset is created—the first publicly available, expertly curated, and pathologically confirmed image dataset with diverse skin tones and identifies important weaknesses and biases in dermatology AI that should be addressed for reliable application to diverse patients and diseases.
Abstract: An estimated 3 billion people lack access to dermatological care globally. Artificial intelligence (AI) may aid in triaging skin diseases and identifying malignancies. However, most AI models have not been assessed on images of diverse skin tones or uncommon diseases. Thus, we created the Diverse Dermatology Images (DDI) dataset—the first publicly available, expertly curated, and pathologically confirmed image dataset with diverse skin tones. We show that state-of-the-art dermatology AI models exhibit substantial limitations on the DDI dataset, particularly on dark skin tones and uncommon diseases. We find that dermatologists, who often label AI datasets, also perform worse on images of dark skin tones and uncommon diseases. Fine-tuning AI models on the DDI images closes the performance gap between light and dark skin tones. These findings identify important weaknesses and biases in dermatology AI that should be addressed for reliable application to diverse patients and diseases.
35 citations
Stanford University1, Instituto Superior Técnico2, University of Queensland3, University of Central Arkansas4, Microsoft5, University of Barcelona6, University of Sydney7, Royal Prince Alfred Hospital8, Emory University9, Memorial Sloan Kettering Cancer Center10, Kitware11, Medical University of Vienna12, National and Kapodistrian University of Athens13, United States Department of Veterans Affairs14, Vanderbilt University Medical Center15, Vanderbilt University16
TL;DR: In this article, the 19 members of the International Skin Imaging Collaboration AI working group volunteered to provide a consensus statement to consolidate limited existing literature with expert opinion to guide developers and reviewers of dermatology AI.
Abstract: Importance The use of artificial intelligence (AI) is accelerating in all aspects of medicine and has the potential to transform clinical care and dermatology workflows. However, to develop image-based algorithms for dermatology applications, comprehensive criteria establishing development and performance evaluation standards are required to ensure product fairness, reliability, and safety. Objective To consolidate limited existing literature with expert opinion to guide developers and reviewers of dermatology AI. Evidence review In this consensus statement, the 19 members of the International Skin Imaging Collaboration AI working group volunteered to provide a consensus statement. A systematic PubMed search was performed of English-language articles published between December 1, 2008, and August 24, 2021, for "artificial intelligence" and "reporting guidelines," as well as other pertinent studies identified by the expert panel. Factors that were viewed as critical to AI development and performance evaluation were included and underwent 2 rounds of electronic discussion to achieve consensus. Findings A checklist of items was developed that outlines best practices of image-based AI development and assessment in dermatology. Conclusions and relevance Clinically effective AI needs to be fair, reliable, and safe; this checklist of best practices will help both developers and reviewers achieve this goal.
34 citations
TL;DR: At present there is no standard nomenclature for describing the diversity of human constitutive skin colour.
Abstract: At present there is no standard nomenclature for describing the diversity of human constitutive skin colour.
11 citations
TL;DR: This paper conducted a scoping review of the literature to address the question, "What are the current and potential impacts of AI technologies on health equity in oncology?"Following the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines for scoping reviews, systematically searched MEDLINE and Embase electronic databases from January 2000 to August 2021 for records engaging with key concepts of AI, health equity, and cancer.
Abstract: The field of oncology is at the forefront of advances in artificial intelligence (AI) in health care, providing an opportunity to examine the early integration of these technologies in clinical research and patient care. Hope that AI will revolutionize health care delivery and improve clinical outcomes has been accompanied by concerns about the impact of these technologies on health equity.We aimed to conduct a scoping review of the literature to address the question, "What are the current and potential impacts of AI technologies on health equity in oncology?"Following PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines for scoping reviews, we systematically searched MEDLINE and Embase electronic databases from January 2000 to August 2021 for records engaging with key concepts of AI, health equity, and oncology. We included all English-language articles that engaged with the 3 key concepts. Articles were analyzed qualitatively for themes pertaining to the influence of AI on health equity in oncology.Of the 14,011 records, 133 (0.95%) identified from our review were included. We identified 3 general themes in the literature: the use of AI to reduce health care disparities (58/133, 43.6%), concerns surrounding AI technologies and bias (16/133, 12.1%), and the use of AI to examine biological and social determinants of health (55/133, 41.4%). A total of 3% (4/133) of articles focused on many of these themes.Our scoping review revealed 3 main themes on the impact of AI on health equity in oncology, which relate to AI's ability to help address health disparities, its potential to mitigate or exacerbate bias, and its capability to help elucidate determinants of health. Gaps in the literature included a lack of discussion of ethical challenges with the application of AI technologies in low- and middle-income countries, lack of discussion of problems of bias in AI algorithms, and a lack of justification for the use of AI technologies over traditional statistical methods to address specific research questions in oncology. Our review highlights a need to address these gaps to ensure a more equitable integration of AI in cancer research and clinical practice. The limitations of our study include its exploratory nature, its focus on oncology as opposed to all health care sectors, and its analysis of solely English-language articles.
6 citations
References
More filters
291 citations
TL;DR: Dermatologist-determined FST is more accurate than self-report for FST III through VI and Rephrasing the questions using specific descriptors that have meaning to people with skin of color may allow physicians to more accurately assign a skin phototype and, by inference, assess the risk of these participants developing skin cancer.
Abstract: Importance Determining Fitzpatrick skin phototypes (FST) allows physicians to assess a person’s risk of developing sunburn and, by extension, the need for sun protection to prevent the development of skin cancer. Reflectance spectrophotometry objectively measures the melanin index and can assist in determining the accuracy of self-report of FST compared with dermatologist-determined FST. Objectives To assess whether self-reported or dermatologist-determined FST is more accurate in identifying a participant’s FST for FST I through VI and to assess the relevance of the burning and tanning measures for a range of skin types among ethnically diverse participants. Design and Setting A convenience sample of participants in an observational study from June 2, 2010, through December 15, 2010, at an ambulatory academic dermatologic practice and employee health center in an urban city. Participants Participants, staff, and students of Northwestern University, who self-identified as being non-Hispanic white, Hispanic or Latino, Asian or Pacific Islander, or black. Main Outcomes and Measures Melanin index as measured with reflectance spectrophotometry compared with dermatologist- and participant-determined FST. Results Forty-two percent (114 of 270) of the participants’ responses to the burning and tanning questions could not be classified using standard FST definitions. The spectrophotometry measurements for dermatologist-determined FST were significantly different for FST III and IV ( P P P P = .90). Participant responses to burning and the dermatologist-determined FST were significantly correlated (Spearman ρ, 0.764; P P = .15). Spectrophotometry measurements assessing FST were statistically significantly different for FST III through VI ( P Conclusions and Relevance Dermatologist-determined FST is more accurate than self-report for FST III through VI. Rephrasing the questions using specific descriptors that have meaning to people with skin of color, such as skin irritation, tenderness, itching, or skin becoming darker, may allow physicians to more accurately assign a skin phototype and, by inference, assess the risk of these participants developing skin cancer. Trial Registration clinicaltrials.gov Identifier:NCT01124513
120 citations
98 citations
TL;DR: The concept of skin type and its relation to skin color, as well as critically appraising the various available methods of skin typing, are discussed.
72 citations
TL;DR: In this paper, a cross-sectional survey collected responses from 3386 individuals regarding self-reported FSPT, pigmentary phenotypes, race, age, and sex, and univariate and multivariate logistic regression analyses were performed to determine variables that significantly predict FSPTs.
Abstract: Background Fitzpatrick skin phototype (FSPT) is the most common method used to assess sunburn risk and is an independent predictor of skin cancer risk. Because of a conventional assumption that FSPT is predictable based on pigmentary phenotypes, physicians frequently estimate FSPT based on patient appearance. Objective We sought to determine the degree to which self-reported race and pigmentary phenotypes are predictive of FSPT in a large, ethnically diverse population. Methods A cross-sectional survey collected responses from 3386 individuals regarding self-reported FSPT, pigmentary phenotypes, race, age, and sex. Univariate and multivariate logistic regression analyses were performed to determine variables that significantly predict FSPT. Results Race, sex, skin color, eye color, and hair color are significant but weak independent predictors of FSPT ( P Limitations Our study enriched for responses from ethnic minorities and does not fully represent the demographics of the US population. Conclusions Patient self-reported race and pigmentary phenotypes are inaccurate predictors of sun sensitivity as defined by FSPT. There are limitations to using patient-reported race and appearance in predicting individual sunburn risk.
68 citations