Over the last two decades, medical practice has advanced at a phenomenal speed. Diagnosis and treatment of a broad range of medical conditions have gone beyond human intervention alone to being assisted and remotely monitored using artificial intelligence (AI)-enabled machines and smart health devices.
Despite rapid advancements in technology and medicine, racial and ethnic health disparities persist in our healthcare system. African Americans remain at a disproportionate risk of chronic health conditions, including diabetes, hypertension, and asthma, all known to worsen Covid-19 outcomes in infected patients. Yet, they represent some of the largest populations of underserved citizens in the U.S. Black men today are dying from preventable diseases due to racial disparities in access to healthcare and socioeconomic barriers. African Americans also have the highest risk of hypertension, a risk factor for cardiovascular disease and stroke. Black people were twice as likely to die from hypertension than whites in 2018 in the same year that the death rate for African American men was 206.6 out of every 100 thousand people.
AI is a broad concept with many separate areas; however, it is fundamentally the technology that allows machines to imitate human behavior. Natural language processing (NLP), for example, allows computers to comprehend human speech. NLP can improve clinical documentation and patient-provider interactions using patients’ electronic health records (EHR). Machine learning (ML), another subset of AI, allows computers to learn and improve from experience without the need for continuous programming. It can help to identify cancerous tissues from images, predict clinical outcomes, and make changes in drug treatment protocols.
Today, wearable glucose monitors utilize AI to monitor blood glucose levels in diabetic patients and send the data to smartphones or wearable devices. Physicians can access patients' electronic health records (EHR) stored securely on the Cloud. Smart inhalers developed by Propeller Health help to track asthma symptoms, make accurate allergen forecasts, and share the reports with physicians to help people with asthma and Chronic Obstructive Pulmonary Disease (COPD). Other emerging applications include diagnosis and monitoring of blood pressure.
A common term in the computing world is "garbage-in, garbage out" (GIGO). What this means for AI is that it's only as great as the data used in training it. If the data capture information from a certain race, e.g., diabetic foot ulcer symptoms in white populations, then it will not do well at identifying the same in dark-skinned people. Similarly, if the AI diagnosis for an eye defect used data that targeted mostly green and blue eyes, it could show errors in analyzing dark brown eyes. Bias can come from a wide range of sources; how you collate the data, whom the study group was, and even personal bias of those involved in the collation. So, the data must be all-inclusive to avoid disparities that can affect the quality of care and overall patient satisfaction.
Moreover, there are technological barriers; some patients do not have a smartphone, computer, or broadband coverage to access digital healthcare, e.g., in rural American communities. In 2019, Chanita Hughes-Halbert, Ph.D., and researchers from the Medical University of South Carolina (MUSC) used natural language processing to find socially-isolated patients from clinical notes. "We know from careful evidence that social determinants are essential to healthcare and health outcomes," she said. "Social isolation is an important social determinant because it reflects the extent to which people perceive they have a high level of connectedness and support," she added.
Artificial intelligence has emerged as an effective tool for promoting equity in the patient-donor relationship and improving healthcare accessibility and quality for everyone, regardless of race, sex, or socioeconomic status. With its video-based mobile app for diabetic wound measurement and real-time remote vitals monitoring, the Black Barbershop is leveraging AI functionality and cloud services to provide access to quality healthcare in underserved Black communities.