Addressing the Challenge of Biomedical Data Inequality: An Artificial Intelligence Perspective

Abstract

Artificial intelligence (AI) and other data-driven technologies hold great promise to transform healthcare and confer the predictive power essential to precision medicine. However, the existing biomedical data, which are a vital resource and foundation for developing medical AI models, do not reflect the diversity of the human population. The low representation in biomedical data has become a significant health risk for non-European populations, and the growing application of AI opens a new pathway for this health risk to manifest and amplify. Here we review the current status of biomedical data inequality and present a conceptual framework for understanding its impacts on machine learning. We also discuss the recent advances in algorithmic interventions for mitigating health disparities arising from biomedical data inequality. Finally, we briefly discuss the newly identified disparity in data quality among ethnic groups and its potential impacts on machine learning. Expected final online publication date for the Annual Review of Biomedical Data Science, Volume 6 is August 2023. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

Optimizing clinico-genomic disease prediction across ancestries: a machine learning strategy with Pareto improvement

“…Therefore, how to utilize data from different ancestry groups is crucial in a multi-ancestral machine learning strategy. We have categorized multi-ancestral (or multi-ethnic) machine learning schemes based on the way they utilize data from different subpopulations [ 7 , 8 ] (Table 2 ). Mixture learning indistinctly uses data from all ancestral populations for model training and testing.…”

“…Recent advances in high-throughput genotyping and genome sequencing technologies have enabled genome-wide association studies (GWAS) in large cohorts, providing the foundation for genomic disease prediction. However, more than 80% of the existing GWAS data were acquired from individuals of European descent [ 1 – 7 ], and the ancestral (or ethnic) diversity in GWAS has not improved in recent years [ 1 , 5 ]. The lack of adequate genomic data for non-European populations, who make up approximately 84% of the world’s population, results in low-quality artificial intelligence (AI) models for these data-disadvantaged populations (DDPs).…”

“…The lack of adequate genomic data for non-European populations, who make up approximately 84% of the world’s population, results in low-quality artificial intelligence (AI) models for these data-disadvantaged populations (DDPs). Genomic data inequality is thus emerging as a significant health risk and a new source of health disparities [ 7 , 8 ].…”

“…We have developed a framework to address the impact of biomedical data inequality and distribution shift on multi-ancestral machine learning and demonstrated its effectiveness in cancer progression and survival prediction tasks [ 7 , 8 , 26 ]. This study extends that framework to optimize clinico-genomic disease prediction across ancestries.…”

Initial study on an expert system for spine diseases screening using inertial measurement unit

“…An important element of our approach is the cyber security of the proposed solution 100,101 . The rapid development of artificial intelligence, including machine learning, is suggesting to us more and more new methods and techniques falling within the framework of computational intelligence that can accelerate and increase the accuracy of analyze, inference and prediction from data [102][103][104] .…”