What does an AI-generated "cancer survivor" look like? An analysis of images generated by text-to-image tools
These videos have been matched automatically. Contact us if they are not relevant.
These videos have been matched automatically. Contact us if they are not relevant.
View abstract on PubMed
Summary
This summary is machine-generated.Artificial intelligence (AI)-generated images of cancer survivors often portray them as young, healthy, and happy, failing to reflect the diverse realities of cancer survivorship. These narrow depictions may reinforce limited societal narratives about cancer recovery and well-being.
Area Of Science
- Digital Health
- Medical Imaging
- Cancer Survivorship Research
Background
- Societal narratives of cancer survivorship frequently emphasize a positive "return-to-normal" post-treatment.
- These dominant representations influence public and survivor perceptions of the cancer journey.
- Artificial intelligence (AI) image generation tools are increasingly used, potentially shaping these narratives.
Purpose Of The Study
- To characterize artificial intelligence (AI)-generated images of cancer survivors.
- To compare AI-generated images of cancer survivors with those of cancer patients.
- To understand how AI images reflect and potentially amplify existing survivorship narratives.
Main Methods
- Generated 160 AI images using DALL-E and Stable Diffusion (40 each for survivors and patients).
- Coded images for perceived demographics, affect, health status, illness markers, and setting.
- Utilized chi-square analyses for quantitative comparisons and qualitative coder insights.
Main Results
- AI-generated cancer survivors were predominantly perceived as White (80%), feminine (80%), young (51%), happy (69%), and healthy (80%).
- Survivor images frequently featured pink (64%), cancer ribbons (35%), and headscarves (51%).
- Compared to patient images, survivor images showed individuals perceived as younger, happier, healthier, and less frequently in medical settings or showing illness markers.
Conclusions
- AI-generated images of cancer survivors do not accurately represent the diverse demographics and experiences within this population.
- These AI outputs risk perpetuating narrow and potentially unrealistic views of cancer survivorship.
- The study highlights the need for more inclusive and representative AI image generation in health contexts.
These videos have been matched automatically. Contact us if they are not relevant.
These videos have been matched automatically. Contact us if they are not relevant.
