Dr. Elaine Nsoesie Leads AIM-AHEAD Program at NIH

Beginning in August 2021, Dr. Elaine Nsoesie, Assistant Director of Research and Faculty Lead for the Racial Data Lab at the BU Center for Antiracist Research, has led the Artificial Intelligence/Machine Learning Consortium to Advance Health Equity and Researcher Diversity (AIM-AHEAD) program at the National Institute of Health (NIH).

AIM-AHEAD is focused on enhancing the participation and representation of researchers and communities who are not typically involved in developing artificial intelligence and machine learning models, due to a current lack of racial, ethnic, and gender diversity in the field. This  program is not a traditional NIH grant, as instead of funding one organization the goal of AIM-AHEAD is to build a consortium, including many organizations and institutions, who will collaborate and implement ideas to (1) increase representation and diversity in AI, and (2) address health disparities and inequities using AI. Ultimately, AIM-AHEAD wants more leaders in this space to be part of the communities that they are developing solutions for.

Dr. Nsoesie has a deep history with data science and artificial intelligence, and has previously worked with organizations such as Data Science Africa, a Kenyan non-profit workshop which brings together African people interested in data science, and Black in AI, an organization started with the goal of increasing the presence and inclusion of Black people in the field of artificial intelligence.

Dr. Nsoesie sees her work at the BU Center for Antiracist Research to be closely aligned with the AIM-AHEAD program: “At the Center we’re thinking about racial equity, and at NIH we’re thinking about health equity which, of course, includes racial equity. There is a lot of crossover and I am learning a lot from both at the same time.”

For instance, Dr. Nsoesie and her colleagues at the Racial Data Lab are part of a broader conversation around racial data and how it is being used. Artificial intelligence and machine learning algorithms are being integrated across every realm of society, including medical fields, which creates an opportunity to use AI to improve quality of care and reduce health disparities. 

However, as Dr. Nsoesie highlights in a recent publication,when biased data is used in those algorithms it can introduce racial discrimination into medical decision making – even when that is not the intention. For example, pulse oximeters are typically calibrated using light-skinned individuals, with the assumption that skin pigment does not matter. These devices do not correctly capture oxygenation levels for darker skin at low oxygen saturation, so the collected  data is biased. Similarly, if an algorithm used for detecting skin diseases is only trained on people with lighter skin, it will  produce less accurate results for people with darker skin, which can lead to serious underdiagnosis. Such discriminatory algorithms have grave effects on individuals and entire populations since they determine who does, or does not, receive needed care. 

We can reduce rather than promote health inequalities by studying and acknowledging historical injustices against marginalized groups and adopting systematic antiracist policies and practices. “Bias is pervasive in clinical devices, interventions, and interactions,” Dr. Nsoesie and her co-author Dr. Marzyeh Ghassemi write, “The solutions to addressing these engrained biases are not easy and require intentional efforts by those who develop algorithms and those who use algorithms including, computer scientists, engineers, clinicians, healthcare institutions, and others.” 

The AIM-AHEAD program is a wonderful opportunity for Dr. Nsoesie, and the new consortium, to begin to create these solutions.