Posters
Presenting Author Academic/Professional Position
Medical Student
Academic Level (Author 1)
Medical Student
Academic Level (Author 2)
Medical Student
Academic Level (Author 3)
Medical Student
Academic Level (Author 4)
Medical Student
Academic Level (Author 5)
Faculty
Presentation Type
Poster
Discipline Track
Community/Public Health
Abstract Type
Research/Clinical
Abstract
Background: Health disparities in dermatological care tend to disproportionately affect underserved populations, particularly those in rural and resource-limited settings. Teledermatology, augmented by artificial intelligence (AI), presents an opportunity to address these inequities by providing underserved populations with improved diagnostic capabilities. Despite its promise, AI models in teledermatology can often exhibit biases, such as reduced diagnostic efficacy for Fitzpatrick skin types IV-VI, due to their training on non-representative datasets. This literature review explores AI's role in teledermatology and potential strategies to enhance its utility for underserved populations.
Methods: A comprehensive review of peer-reviewed literature was conducted to evaluate the integration of AI in teledermatology, with a focus on diagnostic performance, dataset diversity, and accessibility in underserved areas. Emphasis was placed on studies employing convolutional neural networks (CNNs) for image analysis, with additional consideration of community-based initiatives, such as the role of community health workers (CHWs) in mitigating technological barriers. Analytical frameworks included equity-driven evaluations of AI applications and case studies demonstrating the real-world impact of teledermatology.
Results: AI-enhanced teledermatology has demonstrated significant potential to reduce diagnostic delays and improve healthcare access for underserved populations. Studies highlighted high sensitivity and specificity for conditions like melanoma yet persistent gaps in diagnostic accuracy for darker skin tones. Initiatives like the Fitzpatrick 17k dataset and community-driven solutions, such as CHW-led patient education, have shown promise in addressing these challenges. However, infrastructural disparities and ethical concerns regarding AI transparency remain substantial barriers.
Conclusions: AI in teledermatology holds immense potential to address dermatological health disparities, but its impact is contingent upon addressing biases, infrastructural gaps, and ethical concerns. Future efforts must prioritize the development of inclusive datasets, culturally competent algorithms, and equitable technology distribution. Through interdisciplinary collaboration and targeted interventions, AI-driven teledermatology can advance health equity and ensure access to quality care for all populations.
Recommended Citation
Hensley, Jared; Marupudi, Smaran; Martin, Blake; Penmetcha, Neharika; and Villegas, Maria, "Leveraging AI in Teledermatology: A Literature Review on Advancing Health Equity for Underserved Populations" (2025). Research Symposium. 71.
https://scholarworks.utrgv.edu/somrs/2025/posters/71
Included in
Leveraging AI in Teledermatology: A Literature Review on Advancing Health Equity for Underserved Populations
Background: Health disparities in dermatological care tend to disproportionately affect underserved populations, particularly those in rural and resource-limited settings. Teledermatology, augmented by artificial intelligence (AI), presents an opportunity to address these inequities by providing underserved populations with improved diagnostic capabilities. Despite its promise, AI models in teledermatology can often exhibit biases, such as reduced diagnostic efficacy for Fitzpatrick skin types IV-VI, due to their training on non-representative datasets. This literature review explores AI's role in teledermatology and potential strategies to enhance its utility for underserved populations.
Methods: A comprehensive review of peer-reviewed literature was conducted to evaluate the integration of AI in teledermatology, with a focus on diagnostic performance, dataset diversity, and accessibility in underserved areas. Emphasis was placed on studies employing convolutional neural networks (CNNs) for image analysis, with additional consideration of community-based initiatives, such as the role of community health workers (CHWs) in mitigating technological barriers. Analytical frameworks included equity-driven evaluations of AI applications and case studies demonstrating the real-world impact of teledermatology.
Results: AI-enhanced teledermatology has demonstrated significant potential to reduce diagnostic delays and improve healthcare access for underserved populations. Studies highlighted high sensitivity and specificity for conditions like melanoma yet persistent gaps in diagnostic accuracy for darker skin tones. Initiatives like the Fitzpatrick 17k dataset and community-driven solutions, such as CHW-led patient education, have shown promise in addressing these challenges. However, infrastructural disparities and ethical concerns regarding AI transparency remain substantial barriers.
Conclusions: AI in teledermatology holds immense potential to address dermatological health disparities, but its impact is contingent upon addressing biases, infrastructural gaps, and ethical concerns. Future efforts must prioritize the development of inclusive datasets, culturally competent algorithms, and equitable technology distribution. Through interdisciplinary collaboration and targeted interventions, AI-driven teledermatology can advance health equity and ensure access to quality care for all populations.
