Browsing by Subject "racial identity"

Now showing 1 - 6 of 6
  • Thumbnail Image
    Item
    AI recognition of patient race in medical imaging: a modelling study
    (Elsevier, 2022-06) Gichoya, Judy Wawira; Banerjee, Imon; Bhimireddy, Ananth Reddy; Burns, John L.; Celi, Leo Anthony; Chen, Li-Ching; Correa, Ramon; Dullerud, Natalie; Ghassemi, Marzyeh; Huang, Shih-Cheng; Kuo, Po-Chih; Lungren, Matthew P.; Palmer, Lyle J.; Price, Brandon J.; Purkayastha, Saptarshi; Pyrros, Ayis T.; Oakden-Rayner, Lauren; Okechukwu, Chima; Seyyed-Kalantari, Laleh; Trivedi, Hari; Wang, Ryan; Zaiman, Zachary; Zhang, Haoran; BioHealth Informatics, School of Informatics and Computing
    Background Previous studies in medical imaging have shown disparate abilities of artificial intelligence (AI) to detect a person's race, yet there is no known correlation for race on medical imaging that would be obvious to human experts when interpreting the images. We aimed to conduct a comprehensive evaluation of the ability of AI to recognise a patient's racial identity from medical images. Methods Using private (Emory CXR, Emory Chest CT, Emory Cervical Spine, and Emory Mammogram) and public (MIMIC-CXR, CheXpert, National Lung Cancer Screening Trial, RSNA Pulmonary Embolism CT, and Digital Hand Atlas) datasets, we evaluated, first, performance quantification of deep learning models in detecting race from medical images, including the ability of these models to generalise to external environments and across multiple imaging modalities. Second, we assessed possible confounding of anatomic and phenotypic population features by assessing the ability of these hypothesised confounders to detect race in isolation using regression models, and by re-evaluating the deep learning models by testing them on datasets stratified by these hypothesised confounding variables. Last, by exploring the effect of image corruptions on model performance, we investigated the underlying mechanism by which AI models can recognise race. Findings In our study, we show that standard AI deep learning models can be trained to predict race from medical images with high performance across multiple imaging modalities, which was sustained under external validation conditions (x-ray imaging [area under the receiver operating characteristics curve (AUC) range 0·91-0·99], CT chest imaging [0·87-0·96], and mammography [0·81]). We also showed that this detection is not due to proxies or imaging-related surrogate covariates for race (eg, performance of possible confounders: body-mass index [AUC 0·55], disease distribution [0·61], and breast density [0·61]). Finally, we provide evidence to show that the ability of AI deep learning models persisted over all anatomical regions and frequency spectrums of the images, suggesting the efforts to control this behaviour when it is undesirable will be challenging and demand further study. Interpretation The results from our study emphasise that the ability of AI deep learning models to predict self-reported race is itself not the issue of importance. However, our finding that AI can accurately predict self-reported race, even from corrupted, cropped, and noised medical images, often when clinical experts cannot, creates an enormous risk for all model deployments in medical imaging. Funding National Institute of Biomedical Imaging and Bioengineering, MIDRC grant of National Institutes of Health, US National Science Foundation, National Library of Medicine of the National Institutes of Health, and Taiwan Ministry of Science and Technology.
  • Thumbnail Image
    Item
    Dimensions of Black Identity Predict System Justification
    (Sage, 2016-04) Shockley, Ellie; Wynn, Ashley; Ashburn-Nardo, Leslie; Department of Psychology, School of Science
    What explains variability in African Americans’ sociopolitical attitudes? System justification theory implicates both high- and low-status groups in the maintenance of the socioeconomic and political system, postulating that individuals are motivated to justify the system. Self-interest offers a simple explanation for system justification among high-status groups. However, system justification among African Americans is less well-understood. Using a socioeconomically diverse sample of 275 Black undergraduates, including traditional as well as older students, the current survey and quantitative analyses further understanding of attitudes toward the system and institutions by linking attitudes with Black identity. Findings revealed that highly identifying as Black negatively predicted system justification but not if one rejects a Black nationalist ideology. Endorsing an assimilation ideology positively predicted system justification. An oppressed minority ideology did not predict system justification but positively predicted trust across institutions (police and local and national government). Finally, the Black nationalist ideology negatively predicted trust in police. These findings reveal the utility of a multidimensional model of Black identity in shedding light on attitudes toward the system and institutions.
  • Thumbnail Image
    Item
    The Effect of Resume Whitening on African Americans Ingroup Members' Perceived Likability, Hireability, Future Encounters, and Emotional Reactions: The Role of Perceived Racial Identity
    (2021-03) Abdul Karim, Muhammad Fazuan; Ashburn-Nardo, Leslie; Pietri, Evava; Williams, Jane
    Members of stigmatized racial groups who realize that they might face employment discrimination may engage in résumé whitening, whereby they downplay the role of their group identity in their résumés. Although it has been documented that this approach helps members of stigmatized groups, such as Black American and Asian American individuals, move forward in their pursuit of employment (Kang, DeCelles, Tilcsik, & Jun, 2016), little is known about how their ingroup members would perceive this behavior. The current study explores the potential backlash coming from their own ingroup members when Black targets engage in résumé whitening.
  • Thumbnail Image
    Item
    Get involved : stories of the Caribbean postcolonial black middle class and the development of civil society
    (2018-03-07) Williams-Pulfer, Kim N.; Stanfield II., John H.; Springer, Jennifer Thorington; Benjamin, Lehn; Steensland, Brian
    The main research question of this project is: How do the narratives of Caribbean black middle class civil society within the bounds of the “post-postcolonial” state, explain the evolving yet current environment of local and postcolonial civil society development? Using the Bahamas as a case, this project explores the historical, political, cultural, and social conditions that supported the development of civil society within the context of a postcolonial society. Furthermore, an investigation via in-depth interviews, participation observation, archival, and contemporary document analysis contextualizes the present-day work of civil society leaders in the Bahamas. Methodologically, the project employs narrative analysis to uncover the perspectives, voices, and practices of black middle-class Bahamian civil society offering an unfolding, dynamic, and nuanced approach for understanding the historical legacies and contemporary structure of local civil society and philanthropy. The study focuses on three primary forms of narratives. These include the narratives of the past (historical), the narratives of expressive and aesthetic cultural practices, and the narratives of lived experience. The project locates that the development of civil society is linked to historical and cultural forces. The findings show that that the narratives of history, social, and artistic development foregrounds a hybrid model of civil society development drawn from the experience of slavery, colonialism, decolonization, as well as the emerging structures related to economic and political globalization. Furthermore, observed through resilience narratives, local civil society leaders negotiate the boundaries of hybridity in their understanding of their personal, social, and professional identities as well as the way in which they engage government, the public, as well as local and international funders.
  • Thumbnail Image
    Item
    The Promotive and Protective Role of Racial Identity Profiles
    (2020-05) Clifton, Richelle Lee; Zapolski, Tamika C.B.; Pietri, Evava S.; Wu, Wei
    AIM Racial identity has been shown to buffer against the effects of racial discrimination among African Americans. Recently, researchers have developed a more comprehensive assessment of racial identity through the construction of profiles. These profiles help better identify combinations of racial identity that are most protective, as well as those that have the potential to increase risk. To date a majority of the research has been conducted on internalizing and academic outcomes, with limited research on externalizing outcomes, such as substance use. The current study aimed to fill this gap in the literature. METHODS 345 African American college students (80.0% female, 88.4% USA-born, and Mage=21.56) completed measures on racial identity, racial discrimination, internalizing symptomology, academic motivation, and substance use. RESULTS Four racial identity profiles were identified and labeled race-focused (n=228), multiculturalist (n=64), integrationist (n=38), and undifferentiated (n=15). Several direct effects were observed. Multigroup analysis, stratified by profile, revealed several direct relationships between racial identity profiles and outcomes. The probability of being in the multiculturalist profile was negatively associated with depression and stress and positively associated with academic motivation. The probability of being in the race-focused profile was positively associated with cannabis use and the probability of being in the integrationist profile was negatively associated with academic motivation. Being in the undifferentiated profile was not significantly related to any of the outcomes. Two specific moderating effects were also observed; individuals in the integrationist profile were significantly lower in academic motivation as a result of racial discrimination than individuals in the race-focused profile (b=0.10, SE=0.05, p=0.046). Individuals in the integrationist profile were also higher in stress as a result of racial discrimination than individuals in the race-focused profile, however this effect was only trending toward significance (b=-0.14, SE=0.08, p=0.080). CONCLUSION Based on these results, there is evidence for the differential direct and moderating associations of racial identity profiles with various health and behavioral outcomes, such that some appear protective whereas others increase risk. These findings can be used to inform future research related to racial identity and interventions for African Americans experiencing racial discrimination.
  • Thumbnail Image
    Item
    Reading Race: AI Recognises Patient's Racial Identity In Medical Images
    (arXiv, 2021) Banerjee, Imon; Bhimireddy, Ananth Reddy; Burns, John L.; Celi, Leo Anthony; Chen, Li-Ching; Correa, Ramon; Dullerud, Natalie; Ghassemi, Marzyeh; Huang, Shih-Cheng; Kuo, Po-Chih; Lungren, Matthew P.; Palmer, Lyle; Price, Brandon J.; Purkayastha, Saptarshi; Pyrros, Ayis; Oakden-Rayner, Luke; Okechukwu, Chima; Seyyed-Kalantari, Laleh; Trivedi, Hari; Wang, Ryan; Zaiman, Zachary; Zhang, Haoran; Gichoya, Judy W.; BioHealth Informatics, School of Informatics and Computing
    Background: In medical imaging, prior studies have demonstrated disparate AI performance by race, yet there is no known correlation for race on medical imaging that would be obvious to the human expert interpreting the images. Methods: Using private and public datasets we evaluate: A) performance quantification of deep learning models to detect race from medical images, including the ability of these models to generalize to external environments and across multiple imaging modalities, B) assessment of possible confounding anatomic and phenotype population features, such as disease distribution and body habitus as predictors of race, and C) investigation into the underlying mechanism by which AI models can recognize race. Findings: Standard deep learning models can be trained to predict race from medical images with high performance across multiple imaging modalities. Our findings hold under external validation conditions, as well as when models are optimized to perform clinically motivated tasks. We demonstrate this detection is not due to trivial proxies or imaging-related surrogate covariates for race, such as underlying disease distribution. Finally, we show that performance persists over all anatomical regions and frequency spectrum of the images suggesting that mitigation efforts will be challenging and demand further study. Interpretation: We emphasize that model ability to predict self-reported race is itself not the issue of importance. However, our findings that AI can trivially predict self-reported race -- even from corrupted, cropped, and noised medical images -- in a setting where clinical experts cannot, creates an enormous risk for all model deployments in medical imaging: if an AI model secretly used its knowledge of self-reported race to misclassify all Black patients, radiologists would not be able to tell using the same data the model has access to.