Racial Disparities in Skin Tone Representation of Dermatomyositis Rashes: A Systematic Review

What type of medical illustrations best improve understanding of scientific content? A systematic review

The field of visual medical communication has extensively demonstrated the potential benefits of using images to communicate health and scientific concepts. However, it is still unclear which specific types of visuals enhance understanding for particular audiences. The main goal of this study is to identify trends in how style features, namely levels of realism and detail, affect the understanding of medical illustrations by different target audiences. A systematic review was conducted, searching the following scientific literature databases: APA PsycNet, Biomedcentral, DOAJ, EBSCO, ERIC, JSTOR, Pubmed, Science Direct, and Semantic Scholar. Twenty-six original research articles published between 1986 and 2023 were identified, with half of the studies conducted in the USA and South Africa. The current body of evidence supports the idea that simple medical illustrations - both schematic and realistic - may be more easily understood by audiences, particularly lay audiences. Nevertheless, realistic, more detailed illustrations may be more appropriate for individuals with higher health literacy or for medical professionals. This review also compiles practical tips for illustrators and communicators to create effective visual materials, as well as considerations regarding the role of medical illustrators in developing effective medical illustrations.

Impact of ChatGPT and Large Language Models on Radiology Education: Association of Academic Radiology-Radiology Research Alliance Task Force White Paper

Generative artificial intelligence, including large language models (LLMs), holds immense potential to enhance healthcare, medical education, and health research. Recognizing the transformative opportunities and potential risks afforded by LLMs, the Association of Academic Radiology-Radiology Research Alliance convened a task force to explore the promise and pitfalls of using LLMs such as ChatGPT in radiology. This white paper explores the impact of LLMs on radiology education, highlighting their potential to enrich curriculum development, teaching and learning, and learner assessment. Despite these advantages, the implementation of LLMs presents challenges, including limits on accuracy and transparency, the risk of misinformation, data privacy issues, and potential biases, which must be carefully considered. We provide recommendations for the successful integration of LLMs and LLM-based educational tools into radiology education programs, emphasizing assessment of the technological readiness of LLMs for specific use cases, structured planning, regular evaluation, faculty development, increased training opportunities, academic-industry collaboration, and research on best practices for employing LLMs in education.

The Folk Concept of Nursing in Australia: A Decolonising Conceptual Analysis

This article presents a conceptual analysis of the contemporary understanding of NURSING in Australia and proposes strategies for decolonisation. Through historical reflection and the lens of cultural safety and critical race theory, it examines some conditions which make up this concept, including "Florence Nightingale-influenced practices," "intellectual practitioners," and "whiteness in nursing." This analysis aims to identify conditions which we take to be necessary for the folk concept of NURSING to be satisfied and which result in negative outcomes. The article explores why these conditions are plausibly included in this concept and possible objections to their inclusion. These conditions, and subsequently the concept of NURSING, are then critiqued. In this conceptual analysis of NURSING in Australia, we explore three conditions. By critically examining these conditions through the lens of cultural safety and employing decolonising methodologies, the article sheds light on the complex interplay of historical legacies, contemporary practices and potential negative outcomes within the nursing profession. The conclusions drawn propose a shift toward decolonisation, advocating for a cultural safety framework to address historical injustices and highlights possible ways in which one might amend the concept of nursing to be more inclusive of Aboriginal and Torres Strait Islander people. The need for this change is emphasised by the acknowledgement of historical conditions that perpetuated racism and hindered equitable healthcare. Ultimately, the article advocates for a comprehensive decolonisation of the concept of NURSING in Australia, urging the nursing profession to implement cultural safety for the overall well-being of Aboriginal and Torres Strait Islander people. The authors of this article would like to acknowledge the people of the Dharawal and Dharug language group, who are the custodians of the unceded land we have worked on throughout this project. We would also like to acknowledge Aboriginal and Torres Strait Island people nationwide and warn them that some traumatic aspects of Aboriginal and Torres Strait Islander history are mentioned throughout this article. Always was, always will be, Aboriginal land. Two authors on this article identify as Aboriginal, while three do not. Two authors are registered nurses, one is an anthropologist, one is a philosopher and one is a historian.

Ensuring Appropriate Representation in Artificial Intelligence-Generated Medical Imagery: Protocol for a Methodological Approach to Address Skin Tone Bias

BACKGROUND

In medical education, particularly in anatomy and dermatology, generative artificial intelligence (AI) can be used to create customized illustrations. However, the underrepresentation of darker skin tones in medical textbooks and elsewhere, which serve as training data for AI, poses a significant challenge in ensuring diverse and inclusive educational materials.

OBJECTIVE

This study aims to evaluate the extent of skin tone diversity in AI-generated medical images and to test whether the representation of skin tones can be improved by modifying AI prompts to better reflect the demographic makeup of the US population.

METHODS

In total, 2 standard AI models (Dall-E [OpenAI] and Midjourney [Midjourney Inc]) each generated 100 images of people with psoriasis. In addition, a custom model was developed that incorporated a prompt injection aimed at "forcing" the AI (Dall-E 3) to reflect the skin tone distribution of the US population according to the 2012 American National Election Survey. This custom model generated another set of 100 images. The skin tones in these images were assessed by 3 researchers using the New Immigrant Survey skin tone scale, with the median value representing each image. A chi-square goodness of fit analysis compared the skin tone distributions from each set of images to that of the US population.

RESULTS

The standard AI models (Dalle-3 and Midjourney) demonstrated a significant difference between the expected skin tones of the US population and the observed tones in the generated images (P<.001). Both standard AI models overrepresented lighter skin. Conversely, the custom model with the modified prompt yielded a distribution of skin tones that closely matched the expected demographic representation, showing no significant difference (P=.04).

CONCLUSIONS

This study reveals a notable bias in AI-generated medical images, predominantly underrepresenting darker skin tones. This bias can be effectively addressed by modifying AI prompts to incorporate real-life demographic distributions. The findings emphasize the need for conscious efforts in AI development to ensure diverse and representative outputs, particularly in educational and medical contexts. Users of generative AI tools should be aware that these biases exist, and that similar tendencies may also exist in other types of generative AI (eg, large language models) and in other characteristics (eg, sex, gender, culture, and ethnicity). Injecting demographic data into AI prompts may effectively counteract these biases, ensuring a more accurate representation of the general population.

Anatomy's missing faces: An assessment of representation gaps in atlas and textbook imagery

Previous research suggests that underrepresentation in medical curricula perpetuates inequities in healthcare. This study aimed to quantify the prevalence of human phenotypic diversity (e.g., skin tone, sex, body size, and age) across 11 commonly used anatomy atlases and textbooks in pre-clerkship medical education, published from 2015 to 2020. A systematic visual content analysis was conducted on 5001 images in which at least one phenotypic attribute was quantifiable. Anatomy images most prevalently portrayed light skin tones, males, persons with intermediate body sizes, and young to middle-aged adults. Of the 3883 images in which there was a codable skin tone, 81.2% (n = 3154) depicted light, 14.3% (n = 554) depicted intermediate, and 4.5% (n = 175) depicted dark skin tones. Of the 2384 images that could be categorized into a sex binary, 38.4% (n = 915) depicted females and 61.6% (n = 1469) depicted males. A male bias persisted across all whole-body and regional-body images, including those showing sex organs or those showing characteristics commonly associated with a specific sex (e.g. for males, facial hair and/or muscle hypertrophy). Within sex-specific contexts, darker skin was underrepresented, but male depictions displayed greater overall skin tone variation. Although most images could not be assigned to a body size or age category, when codable, these images overwhelmingly depicted adults (85.0%; 482 of 567) with smaller (34.7%; 93 of 268) or intermediate (64.6%; 173 of 268) body sizes. Ultimately, these outcomes provide reference metrics for monitoring ongoing and future efforts to address representation inequalities portrayed in anatomical imagery.

Racial Disparities in Diagnosis and Treatment of Patients With Dermatomyositis of Different Skin Tones

BACKGROUND

Delays in the diagnosis and treatment of dermatological conditions in minorities are a well-documented health disparity. We aimed to determine if there was a delay in detection and treatment initiation for dermatomyositis (DM) and amyopathic dermatomyositis (ADM) in patients of different skin tones.

METHODS

Patients from Montefiore Medical Center who met the criteria for DM and ADM were included in this cohort study. Records were reviewed for date of first documented rash, creatine kinase levels, muscle weakness complaints, and date of first steroid or disease-modifying antirheumatic drug initiation. The median number of days between rash documentation and therapy initiation was compared for patients of different races, including non-Hispanic White, non-Hispanic Black, Hispanic, and other (Asian and unknown). Data were compared in White versus non-White skin.

RESULTS

Sixty-three DM and 9 ADM patients met the inclusion criteria. There was a shorter time to treatment initiation in White versus non-White patients, with a median number of 8 days compared with 21 days, respectively ( p = 0.05). Kaplan-Meier curves showed prolonged time to diagnosis and treatment in all other races when compared with White patients ( p = 0.03).

DISCUSSION

It took clinicians longer to diagnose and treat DM and ADM in patients of color. The trends observed emphasize the importance of increasing dermatology education of non-White skin to improve detection and treatment of DM and ADM and minimize health disparities.

Visual Misrepresentations: The Lack of Skin Tone and Sex Equity in General Surgical Textbooks

OBJECTIVE

Lack of racial and ethnic diversity in educational material contributes to health disparities. This study sought to determine if images of skin color and sex in general surgery textbooks were reflective of the U.S.

POPULATION

DESIGN

All human figures with discernable sex characteristics and/or skin tone were evaluated independently by 4 coders. Each image was categorized as male or female. Skin tone in each image was categorized using the Massey- Martin skin color scale. This data was compared to 2020 U.S. Census Data.

SETTING

U.S. Medical School.

PARTICIPANTS

Not applicable.

RESULTS

A total of 1179 images were evaluated for skin tone alone; 293 images for sex alone. 650 images depicted characteristics of both sex and skin tone. Interrater reliability was 0.78 for skin tone and 0.91 for sex. While the U.S. population is 59.3% white, 29.5% non-black persons of color and 13.6% black, in surgical textbooks, 90.7% of images were white, 6.5% were non-black persons of color, and 2.8% were black. Distribution of skin tone for all textbooks were significantly different. (p < 0.001) compared to the U.S.

POPULATION

The U.S. population is 49.5% male and 50.5% female. When images of sex specific genitalia and breasts are excluded, surgical textbook images are 62.9% male and 37.1% female. Only 1 textbook had a distribution of sex that was similar to the U.S.

POPULATION

CONCLUSIONS

Despite increasing diversity in the U.S. population there is a lack of skin tone and sex diversity in traditional surgical textbooks.

Skin of color in dermatopathology: does color matter?

Skin of color (SoC) remains an understudied and under taught area of dermatology despite its rising importance. Race and ethnicity play a particularly important role in dermatology as skin pigmentation can affect the manifestation and presentation of many common dermatoses. With this review, we seek to review pertinent differences in SoC histology, as well as highlight the histopathology of conditions more common in SoC and address inherent bias that may affect accurate dermatopathology sign out.

Diagnosing Disorders of Facial Erythema

Identification of specific patterns, shades, and intensity of erythema in the skin has always been critical and fundamental to diagnostic accuracy in dermatology. Erythema is often less noticeable in darker skin types. The interplay of inflammation and variance of skin tone contributes to appreciable differences in the clinical appearance of cutaneous disease in darker complexions. In this article, we discuss common disorders that present with facial erythema in skin of color and offer distinguishing features of each disorder to assist the clinician with diagnosing these conditions in the presence of deeply pigmented skin.

Collagen Vascular Diseases: A Review of Cutaneous and Systemic Lupus Erythematosus, Dermatomyositis, and Distinguishing Features in Skin of Color

Collagen vascular diseases such as lupus erythematosus and dermatomyositis (DM) occur 2 to 3 times more often among patients with skin of color. In this article, the authors review DM and cutaneous lupus erythematosus, including acute cutaneous lupus erythematosus, subacute cutaneous lupus erythematosus, and discoid lupus erythematosus. They discuss the distinguishing features between these entities and highlight distinct presentations and management considerations in patients with skin of color to aid in prompt and correct diagnoses in this patient population.

Racial Differences in the Diagnosis of Lyme Disease in Children

Black children with Lyme disease compared with children of other races were less likely to have an erythema migrans lesion diagnosed (adjusted odds ratio, 0.34; 95% confidence interval, .14-.79) but more likely to have a swollen joint (adjusted odds ratio, 3.68; 95% confidence interval, 2.13-6.36) after adjustment for age and local Lyme incidence.

Assessment and Identification of Skin Disorders in Skin of Color: An Integrative Review

Skin assessment in patients with dark skin tones (DST) continues to be a challenge for many healthcare providers (HCP) because the visual cues are not always readily identified. For example, identification of early signs of pressure injury when subtle skin color changes are missed has the potential to cause harm and contribute to healthcare disparities. Appropriate wound management can begin only when the wound is correctly identified. For HCPs to identify early signs of skin conditions in DST patients, they must be provided education and effective tools enabling them to identify clinically relevant signs of skin damage in all patients. This article reviews basic anatomy of the skin; it focuses on differences seen in DST and reviews assessment strategies to assist the HCP to identify skin changes and conditions.