How AI can Combat Medical Bias in the Health Tech Industry

AI Nexus - How AI can Combat Medical Bias in the Health Tech IndustryImage | AdobeStock.com

Artificial Intelligence (AI) is a complex and diverse field, yet with continued diversification and innovation, it is starting to simplify healthcare. Providing technological solutions to anatomical issues will be crucial in obtaining a sustainable trajectory for worldwide healthcare systems. However, for this to be truly sustainable, we must ensure that these solutions are inclusive.

Traditionally, medical innovation has been designed, calibrated and tested on majority Caucasian western civilisations, which means certain technologies that we consider an everyday staple in healthcare, are less effective when used on darker skin pigmentation or across other ethnic differences.

A prime example is the Oximeter. A study in 2021 by the University of Michigan showed that Black patients were more likely to have low oxygen levels that were missed by pulse oximeters, translating into as many as 1 in 10 inaccurate readings. This represents a fundamental flaw in health technology and ultimately, could result in misdiagnosis or the inability to detect serious health issues.

So what is the solution?

Medical trials overseas

One of the most simple, yet most effective, solutions is to test your technology on a variety of ethnic groups. Recently at AI Nexus Healthcare, we conducted clinical trials in India, including a low income community in TC Palya, Bangalore.

Our motivation behind this was to ensure mia Vitals, our app that provides a health scan from the smartphone camera was able to yield accurate results irrespective of skin pigmentation or religious symbols.

Typically, those with a darker skin pigmentation are more likely to receive an inaccurate reading on traditional technology such as oximeters, however, with our fingertip scanning technology, we were able to yield a 95% test accuracy following two phases of testing SpO2.

As well as this, the face scanning function of the app now functions more accurately in low intensity lighting and is not impacted by traditional and religious symbols such as a bindi, tika or vibhuti.

So what is the takeaway lesson here?

Inclusive digital healthcare is only obtained through inclusive testing

To put it simply, if you want your technology to be effective across a wide range of ethnic groups, then the logical solution is to test its effectiveness on a wide range of ethnic groups. Global healthcare systems are currently unsustainable, with clinicians across the globe under immense pressure and suffering at the hands of excess demand.

Cognitive and explainable AI, essentially AI that can emulate the diagnostic approach of a clinician and explain its reasoning to the user, has the unique ability to alleviate this pressure and support medical professionals. However, populations are growing rapidly across the globe and in many cases, this growth occurs more intensely in developing nations, who lack the medical resources to meet this demand.

Therefore, as technology innovators who want to make a real difference in the world, we need to look beyond single cultures and ethnicities and cater for diverse populations both within the US and overseas.

Why it’s important to prioritise accessibility  

At AI Nexus Healthcare, our technology can monitor vital signs and offer actionable medical advice, which essentially could determine whether medical intervention is needed. Not only does this prevent unnecessary trips to the doctor, it also offers a cost effective, simple and most importantly, accessible insight into the human body.

Ultimately, technology is taking healthcare to new heights each and every day and to truly get the most out of this innovation, inclusivity and accessibility must be a priority. We are all part of a global healthcare community and to neglect any demographic is a fundamental failing. Diversity should always be celebrated and this does not exclude healthcare.

Diversify your testing, refine your product and make a real difference.

 

About the author

Faramarz Farhoodi, founder of AI Nexus Healthcare, has more than three decades of experience in the field of AI, having led more than 1,000 person-years of AI application development across the defense, healthcare, manufacturing and finance industries – with his experience as diverse as helping to build NATO command and control systems to overseeing Amazon.com’s largest commercial account.

 With 80% of chronic disease preventable and a global shortage of healthcare professionals, Faramarz and AI Nexus Healthcare are using a specialized AI skillset to transform healthcare systems, focusing on intervention and prevention.