Nicholas Christakis

Quick summary of our Duke BIG IDEAs Lab paper out today in The Lancet Digital Health: https://bit.ly/3BBAdtj * What did we do?  We analyzed demographic reporting in the 175 human studies on the PhysioNet open-access database. We looked at the extent to which race, ethnicity, sex or gender, and age were reported & how reporting practices vary by geography and study/data type. * What did we find? We found that <7% of studies reported all four key demographic variables, while 14.3% did not report *any* demographic information. Race and ethnicity were reported much less frequently than age and sex or gender. * Why does it matter? This lack of demographic reporting could contribute to algorithmic bias in #AI in #healthcare, which may disproportionately affect marginalized groups and could lead to inequitable healthcare outcomes. * What next? This work highlights the need for standardized reporting practices to ensure the fairness and generalizability of AI-driven healthcare tools. It was a pleasure doing this with our wonderful co-author team: Sarah Jiang (Duke University CS Undergrad), Perisa Ashar (Duke University Pratt School of Engineering BME Undergrad), & Md Mobashir Hasan Shandhi (former postdoc, current Arizona State University faculty).

81

11 Comments

Professor Markus Buehler and postdoc Alireza Ghafarollahi developed AI frameworks to identify evidence-driven hypotheses that could advance biologically inspired materials in new research study. https://lnkd.in/eN2nWW7d

36

At the UN High Level Political Forum in New York, I had the chance to intervene in the "Science, technology and innovation: Triggering transformation and sustaining science-driven solutions" session where I discussed about: ↪ Discrimination based on Work and Descent ♀ The growing digital divide, affecting especially women and girls from our communities vs. the role that inclusive technological advancements play in dismantling systemic barriers. 📶 And emphasized the importance of academic research on our communities for achieving the SDGs. 📌 More: https://lnkd.in/eNf6Gc5B

34

3 Comments

New release: Part 2 of the new R Street Institute “AI and Public Health Series” addresses “How #AI Can Help Tackle Major Causes of Suffering and Death.” This essay is jam-packed with examples and data highlighting the many ways medical professionals are already tapping the power of AI to significantly improve lives. The essay examines how AI and #machinelearning are currently being used to address heart disease, cancers, sepsis and superbugs, strokes / brain disease, paralysis & disabilities, mental health, organ donation, eye diseases, and more. Amazing things are happening thanks to the power of algorithmic technologies and advanced #datascience. https://lnkd.in/egTGqwMz #artificialintelligence #healthcare #medicine #techpolicy #robotics

20

2 Comments

An analysis by the Deloitte Health Equity Institute and the School of Global Health at Meharry Medical College has found that mental health inequities could lead to $14 trillion in excess costs between now and 2040, if left unaddressed. It’s imperative that we increase access to culturally-informed, evidence-based mental healthcare access by reducing stigma and addressing discrimination.   https://lnkd.in/eSa4WUsh #healthequity #mentalhealth

42

4 Comments

The Mount Sinai Health System and IBM Research are joining together to improve mental healthcare via #ArtificialIntelligence and behavioral data. Learn more about the PREDiCTOR study, which received a $20 million grant from the National Institute of Mental Health (NIMH): https://lnkd.in/e3QZchp6

15

1 Comment

I want to recommend this useful and timely report from the Nuffield Council on Bioethics. AI-powered genomic health prediction (AIGHP) is set to transform the future of healthcare by using AI to analyze genetic data and predict disease risks and drug responses. While it promises personalized treatment and preventive care, challenges such as data privacy, genomic discrimination, and uneven accuracy, especially for non-European populations, suggests the need for cautious implementation. Policymakers will need to balance the potential benefits with these risks, ensuring robust safeguards and equitable access. However, the risks can be managed with proper regulation and strategic use. When properly applied, AIGHP should position us to significantly enhance healthcare outcomes and efficiency. #AI #Genomics #Healthcare #DataPrivacy #Innovation

8

This paper by Mlandu et al. 2024 describes evidence-based best practices for inclusive EEG research among infants with tight and curly/afro-textured hair. Super excited to see this research being done in Africa and to read the practical recommendations for scientists to follow when engaged in this research. Congratulations again to the authors! 👏🏾 Please give it a read👌🏾 #neuroscience #inclusivity #eeg #neuroethics

3

My article on the cultural history of durian appears in the journal Philosophies: https://lnkd.in/eFRjfkpY

5

1 Comment

For those interested in AI and Qualitative research. Important discussion. Well worth a listen. 

7

"A new study by MIT researchers examines summertime heat exposure in prisons across the United States and identifies characteristics within prison facilities that can further contribute to a population’s vulnerability to summer heat. The study’s authors used high-spatial-resolution air temperature data to determine the daily average outdoor temperature for each of 1,614 prisons in the U.S., for every summer between the years 1990 and 2023. They found that the prisons that are exposed to the most extreme heat are located in the southwestern U.S., while prisons with the biggest changes in summertime heat, compared to the historical record, are in the Pacific Northwest, the Northeast, and parts of the Midwest." #InsightstoAction #ClimateChange #ClimateJustice https://lnkd.in/eSajgKse

1

1 Comment

Sharing a nice commentary about some recent work from our collaboration with Wilbur A. Lam's group (first author Christina Caruso, Xiaopo Cheng did the computations) on how aberrant red blood cells in the circulation lead to chronic microvascular dysfunction in sickle cell disease: https://lnkd.in/gG5Q6EMQ Here's the paper: https://lnkd.in/gFxRqzR7 #redbloodcells #sicklecelldisease #everythingflows

38

Donors Want Real-World Impact: What It Means for Science in Higher Education Op-ed in Times Higher Education by Prof. Shai Shen-Orr, director of the Zimin Institute for AI in Healthcare Solutions As donations shift toward real-world impact, universities can tap into this trend by emphasizing translational research. Donors today seek to fund projects that address global challenges, from healthcare to environmental issues. Universities reliant on philanthropy, such as in the US and Israel, are adapting by supporting academic spin-offs and developing programs that bridge research with practical applications. Partnerships, like those at the Technion’s Zimin Institute for AI in Healthcare, are key. They bring together donors, researchers, and industry to accelerate innovation. For continued funding, universities must work closely with donors, ensuring clear goals, regular updates, and a shared vision for turning ideas into solutions that make a difference. Read more here: https://lnkd.in/g7Kgp4FD Times Higher Education Zimin Institutes Zimin Foundation

58

2 Comments

Goodes Prize Honoree and DxA-funded researcher, Rhoda Au, PhD, MBA, and her colleagues at Boston University recently published a study on an AI model they developed that can predict if someone with mild cognitive impairment (#MCI) is likely to develop Alzheimer's in the next six years with 78.5% accuracy. The ADDF has long supported Dr. Au's pioneering efforts in developing the first #digital biomarkers for #Alzheimers. Digital biomarkers will be an important component of the future of Alzheimer’s care, helping monitor risk and supporting early-intervention and passive disease detection. Dr. Au’s co-authors also note they hope to develop an #AI-powered smartphone #app to diagnose the disease. Our #SpeechDx initiative, which intends to create the largest repository of #speech and #voice data, hopes to ultimately provide researchers with the high-quality data they need to develop accurate #DigitalTools like this that can accelerate the detection, diagnosis and monitoring of Alzheimer’s. Read more about the study here: https://bit.ly/4eC3gfH

56

2 Comments

INTELLECTUAL EXCHANGE AND GREAT BARBECUE AT NIH(NHLBI)-FUNDED SHORT COURSE ON STRENGTHENING CAUSAL INFERENCE IN BEHAVIORAL OBESITY RESEARCH In July, my colleagues and I offered our annual short course on Strengthening Causal Inference in Behavioral Obesity Research, funded by the NHLBI ( https://lnkd.in/gKx8jb3w ). The course, for which I serve as PI with my co-PIs—Drs. Andrew Brown, Kevin Fontaine, and Colby Vorland—was graciously hosted by the U. of Arkansas for Medical Sciences, Little Rock and featured outstanding speakers. The 5-day course also included time for open discussion, allowing exploration of many topics. Discussions continued into our evening times together. One evening, the faculty all went out for dinner at a fantastic barbecue restaurant ( https://lnkd.in/g2TCQhyQ  ) that was recommended by my friend and colleague, Mario Ferruzzi. While we ate, we discussed the question of when we are able to draw conclusions based on observations, and when it is not reasonable to do so. Drew Sayer then introduced questions about sequential and adaptive designs for assessing treatment response heterogeneity (TRH) and whether good criteria exist for evaluating the plausible magnitude of TRH early in RCTs to decide how to conduct and analyze the remainer of the RCT. We also debated the health effects of eating red meat (while we ate red meat), how well established our knowledge of these effects is, and whether credible evidence exists that red meat is better or worse than other food. And if such evidence does not exist, why do people so strongly take opposite positions on the issue? We further discussed the idea of trying to separate the strength of the evidence for forming a scientific conclusion from the strength of a decision to take (or not take) some action (c.f: https://lnkd.in/gJ_hRzG9 ). In other words, people are sometimes reasonably convinced to take some action on the grounds that it seems plausibly beneficial and this in no way should be equated with a statement that the scientific evidence has, to a reasonable degree of certainty, *demonstrated* that the plausible benefit will occur. From there, we talked about how to grade evidence, a topic on which short course attendee Dr. Bradley Johnston ( https://lnkd.in/gqdXpHDp ) is an expert, and which became a substantial part of the discussion. After a lively conversation, we ended the night with ice cream sundaes at Loblolly Creamery. I express my gratitude to Colby Vorland, IU research scientist, and Cynthia Herrera Alley, our department administrator and program manager, for helping to organize the short course so well. My thanks also to all the speakers and participants who helped make the event a success! Some who may wish to comment: Greg Pavela, Michelle M Bohan Brown PhD, RDN, Aaron Caldwell, Christopher K. Haddock, Ph.D., PStat®, FTOS.

72

1 Comment

I am very eager to read this special issue on AI "in supporting the health, well-being, independence, and quality of life of older adults and their families, as well as the broader impacts AI may have on gerontology" from the Gerontologist. Check out information below on submissions to this special issue that is sure to be of high interest to so many. Joseph Gaugler Gerontological Society of America (GSA) University of Minnesota School of Public Health Minnesota Gerontological Society Center for Healthy Aging and Innovation #aging #AI #artificialintelligence #gerontology

8

Do you know VCU's Egyptian Building? A national landmark, still in use by the Medical School, yet with a dark history rarely discussed. Here's my new blog post about it: https://lnkd.in/e43vB8pQ

6

7 Comments

Alda Center's Elizabeth Bojsza and Stony Brook University School of Communication and Journalism's Xia Zheng have a new article out with Association for the Study of Medical Education (ASME) on the Addressing Distress in Applied Improvisation Participants Tool (ADAPT), a first-of-its-kind protocol for addressing participant distress. The ADAPT can be customized to other applied improv curricula and experiential learning activities. This tool guides facilitators through four phases of attending to participant distress during applied improvisation sessions: Redirecting the Spotlight, Responding to Distress, Reintegrating the Group and Reflecting on the Distress. Check it out here: https://lnkd.in/dUuA9wDb

8