Artificial intelligence promises to revolutionize healthcare. Potential benefits include improving diagnostics, discovering new treatment options, and facilitating personalized medicine.
But what if the AI that powers the data is inaccurate because it relies on studies that aren’t representative of diverse populations? John D. Carpten, Ph.D., distinguished genomics researcher and scientific director of a cancer research and treatment institution City of hopehas these concerns.
“While the mapping of the human genome has ushered in a new era of precision medicine, approximately 95% of data from genomic studies conducted over the past two decades comes from the genomes of white Europeans. Just 3% comes from Asians, and less than 1% are from African Americans or Afro-Caribbeans, Africans and the Latinx community,” says Dr. Carpten.
This is a situation that can lead to underrepresented populations not benefiting from the targeted treatments that precision medicine aims to provide. And medical biases could persist for years.
Scientists have emphasized that clinical studies need to be more inclusive to better understand health disparities. In the United States, African-American men are 76% more likely to be diagnosed with prostate cancer – and 120% more likely to die from it – than white men, according to the National Institutes of Health. However, this phenomenon is only just beginning to be studied.
Multiple myeloma provides another example of a cancer outcome. Although African Americans make up about 14% of the U.S. population, they account for about 20% of multiple myeloma cases.
AI needs diverse data
How do these results limit AI? Most of the data used to generate cancer tumor biomarkers come from populations largely of European origin – although molecular-level differences in tumors are observed in individuals from different backgrounds. If this data is used by AI to extrapolate results, scientists could instead aggravate and perpetuate biases and introduce misinformation, which could exacerbate treatment gaps between groups, as well as between genders and genders. stages of life.
For AI to benefit biomedical research, clinical databases must include and reflect more patient populations. By incorporating as much diversity into the system as possible, results are more likely to be accurate and equitable, reflecting the diverse health care needs of all communities.
Dr. Carpten says that including more diverse patient populations in clinical trials will lead to better health outcomes, consistent with real-world medicine. This approach also improves the accuracy and reliability of AI in healthcare because it ensures that the data used to train AI models is more representative of the broader patient population.
Also address workforce access and diversity
And there are other challenges to overcome. Even if actionable cancer tumor mutations specific to patient populations are identified, the findings will have little impact on patient outcomes unless access to specialized care for these populations also improves.
Dr. Carpten says it’s also important to build a diverse medical workforce, especially in cancer research. “Having scientists and outreach partners who reflect and understand the cultural nuances of the communities they serve helps build trust, foster inclusion, and increase participation in clinical trials,” he says.
“We will get there, one step at a time,” says Dr. Carpten. “As Rev. Jesse Jackson said, “When everyone is included, everyone wins.” »
**This article is for informational purposes only and is not a substitute for professional medical advice. If you are seeking medical advice, diagnosis or treatment, please consult a medical professional or health care provider.