The introduction of artificial intelligence (AI) into health care is revolutionizing the delivery of care. Right now, hospitals are buying AI systems, not with the intention of replacing humans, but to improve care or streamline the administration process. This approach has proved to be particularly helpful in the dire times of the COVID-19 pandemic. AI has helped scientists develop vaccines at an incredible speed when it was most needed, and was used by epidemiologists to track the virus progress.
However, as AI and machine learning systems are becoming better than humans at detecting diseases and costing less, many are legitimately questioning whether it is ethical to replace some healthcare workers.
Some of the newer AI-powered software used to scan medical imaging reports are able to spot details that human eyes cannot find, potentially saving more lives than even the best doctor. More importantly, they can scour the reports for signs of other conditions which may differ from the one that the pathologist was looking for when the test was performed.
They could even be used retroactively to re-scan millions of electronic medical records to find any symptom of an otherwise unknown disease in a fraction of the time required by a human doctor. There’s no need to explain how useful this could be to find otherwise untraceable signs of early infections as early as possible in a hypothetical future pandemic.
In the past, some studies have proven that when an urgent diagnosis is required, a deep learning algorithm is better and quicker at diagnosing cancers than human radiologists. Machines operate better than humans when under pressure, and in most real-world settings they can outperform them since they never get distracted or get tired.
The hardest lesson that the coronavirus pandemic taught us, is that we cannot force an overwhelming amount of infected patients on a small number of health care providers. AI can be central in helping doctors maintain connection to their patients without exposing them to the risk of infection, and ease their burden, saving them from overexertion.
AI is also better at predicting health events and determining which data is relevant to treat a specific patient's disease. Machines can scan through thousands of clinical papers and medical reports in the blink of an eye, and never get overwhelmed by the excess of data. However, even if they can provide useful insights to humans, a real doctor's experience and ability to devise new treatment strategies are still critical.
While humans will always be required to work side by side with machines, there's a chance that some specific medical professions such as the radiologists and pathologists may be replaced. At the least, aspects of their jobs could be supplemented by AI. As more lives could be saved, it could even be unethical not to do so in the near future.