Artificial intelligence (AI) used correctly has big digital business pay off, however, healthcare has been slow to adopt it. Created and programmed to perform and learn intellectual tasks that any human can do, AI may seem like the perfect solution to the world's problems, but it's riddled with its own.
While AI has pervaded most industries, healthcare remains to be one of the slowest sectors to adopt the technology. Unsurprisingly, it's due to notable regulatory limitations, ethical matters, as well as risks associated with patient safety. That’s in addition to the infancy of the AI itself.
In 2018, Robert Challen, M.D., a researcher at the University of Exeter co-authored a study in the UK tackling the key obstacles healthcare has to overcome to integrate AI.
Dr. Challen outlines that the AI research community and healthcare professionals aren’t communicating with each other, resulting in naive assumptions from tech developers on how products will be useful in a real healthcare setting rather than a hypothetical one.
Of course, then there is the obstacle of achieving approval from the US FDA (Food and Drug Administration). While the FDA has taken steps to keep up with digital advancements, it is prudent in approving software that falls under the agency's high-risk category. Clinicians will be apprehensive in adopting software that hasn’t been endorsed by the FDA.
In addition to regulatory hindrances, Dr. Challen flags AI’s biggest drawback is that it’s a black box. Which means that there is no way to tell how algorithms in AI are used to arrive at a particular conclusion. So, if a negative outcome occurs there are issues with tracking an AI application's process to see what specific actions were wrong. This is perhaps the biggest shortcoming, as it doesn't allow users to prevent these mistakes in the future. Also, AI is vulnerable to being fed incorrect data causing misdiagnoses and incorrect treatment medications. These flaws make physicians reluctant to trust the software at hand.
In February, PWC released a report declaring that the healthcare industry has the biggest potential of all to be disrupted by AI in the coming years. While the healthcare sector certainly has its concerns about privacy, protection of safety, and ethics, AI isn't going anywhere. It's only a matter of time the software gains the acceptance of patients, providers, and regulators.