Using artificial intelligence to predict individual health outcomes

One of the most intriguing possibilities of artificial intelligence in medicine is the possibility that artificial intelligence can predict patient outcomes. I would like to discuss three such recent developments.

My first example is the suggestion by health policy experts Michael Millenson and Jennifer Goldsack that AI systems like ChatGPT and Google’s Bard may soon be able to match data from different sources and let patients know which doctors and hospitals can provide the best odds. about treatment success for their specific problems: “You can identify the surgeon in Chicago who does the most knee replacements and his infection rate, find the survival data for breast cancer patients at a prominent medical center in Los Angeles, or get referrals for cardiac surgeons in New York. “

Of course, such recommendations would only be as good as the reliability of the underlying data on patient survival rates or physician complication rates. Also, the frequency of complications alone does not necessarily reflect the competence of the physician. A doctor who sees sicker patients or a surgeon who deals with more challenging cases may have a higher complication rate than others who do not. As always, the quality and context of these data should be factored into any recommendations made to patients. But overall, I think this would be a good development, and I hope that doctors and hospitals will be willing to publish success rates (and complication rates) as transparently as possible.

A second example is the surprising discovery that AI algorithms can detect people with type 2 diabetes by analyzing their voice – sometimes even before the patients or their doctors realize it. In men, there were subtle changes in voice intensity and amplitude; in females there were subtle differences in height variation. The accuracy of computer predictions is approximately 86% for men and 89% for women. (Please see the full paper for a more detailed discussion of specificity, sensitivity, and other statistical results.)

The exact mechanism of this detection method is not yet fully understood. However, the researchers suspect that early diabetes may affect the mechanical properties of patients’ vocal cords and the patients’ ability to control their vocal muscles—changes that can be detected in voice recordings. Although this research is still preliminary, if further studies confirm this result, doctors will have a new inexpensive and non-invasive way to screen for a disease that affects millions of Americans.

My third example of using AI to predict health outcomes comes from work at Vanderbilt University and the University of Missouri–Kansas City. Researchers at these institutions have shown that chest CT data used to screen for early lung cancer can not only be used to predict mortality from lung cancer, but also from heart -vascular diseases or mortality from any kind reason.

Lead researcher Kaiwen Xu explained that this research could help doctors better determine which patients would benefit from interventions such as physical training or lifestyle changes, even at a very early stage before the onset of the disease.

AI is not (yet) a reliable tool for predicting short-term mortality in an emergency situation. But there are researchers looking at the ethical implications of algorithms that could someday make reliable predictions about whether a patient will die within the next 30-60 days, and how that could change the delivery of care in emergency rooms, hospices, etc. n.

A recent survey found that doctors, nurses and healthcare administrators recognize that AI holds great promise for helping to guide appropriate care for people at high risk of mortality, but they would be horrified if “AI technologies were only introduced to save money’.

There’s an old joke usually attributed to Yogi Berra: “It’s hard to make predictions, especially about the future.” But as AIs continue to get better at making both long- and short-term predictions about outcomes for patient health, both patients and physicians will face new opportunities – and challenges – in how best to use this knowledge.

Take a look my website.

Leave a Comment

Your email address will not be published. Required fields are marked *