Imagine the first sound you hear in the morning is your smartphone not only waking you up, but then listing possible breakfast foods that would help lower your cholesterol based on your most recent readings.
Those readings would be obtained by a little prick of the finger, then putting the blood drop on a test strip, and inserting the strip into a special device you connect to your phone that gives you a read-out of your blood-sugar count, cholesterol level and other readings.
The app can monitor your sleep habits, with a smartwatch attached, can calculate your calories burned with exercise, and then recommend a meal plan based on the foods you scanned when you bought them at the store.
It’s coming soon. Health applications for smartphones already exist. Numbers of steps walked, calories burned, hours of sleep – these can all be measured right now. But it’s more information than most doctors have time to look at.
Your smartphone may soon be a key component to monitoring your health.
How can all of this data be analyzed and become useful in medical care? The answer might just lie in something not human at all.
Artificial Intelligence could actually replace your primary care provider and maybe, for some things, it should!
The main causes of death, year after year, are cancer and atherosclerotic cardiovascular disease – heart attacks and strokes. The way to prevent these is to do routine cancer screenings and also monitor the risk factors. But seeing a doctor once or twice a year isn’t the same as getting a daily reminder of progress.
Since up to 95 percent of cancers have been recently found to be associated with lifestyle choices, what if there was an application that could effectively tailor a dietary and exercise program to significantly reduce the risks of both cancer and heart disease, and continue to personalize the advice based on monitoring?
The potential of AI has been underestimated in the medical profession. IBM’s Watson has been “learning” for years now, and it’s not alone. Several companies have entered the arena, all hoping to provide the advanced power of analyzing vast amounts of data and identifying patterns that humans, often, cannot.
Researchers at Google have trained computers to recognize breast cancer cells on microscopic specimen images. They have also increased the ability to detect diabetes changes in the eyes by creating a computerized algorithm to “read” the pictures and diagnose the condition in people who haven’t yet seeng an eye doctor in person. AI has been trained to read x-rays, detect TB — the possibilities seem endless.
One of the biggest areas of potential use is in disease prevention. In the United Kingdom, researchers gave data on almost 300,000 patients to a computer learning algorithms to predict heart attacks in patients and determine risk factors. An additional 80,000 charts were analyzed with predictions compared from the machine’s learning and separately by the physician-written guidelines. Predicted rates were compared to actual rates of heart attacks. Who did better? The machine.
So what can modern doctors do if their best guess might soon be replaced by a smarter computer-based diagnosis?
Well, if you can’t beat them …
The personal touch of doctors will always be needed, but artificial intelligence could help them do their jobs better.
The U.S. health care system is increasingly complex, from the way medical care is provided to the various types of medical record systems being used to the cost of providing all of the services, especially in rural areas.
If computers with artificial intelligence can be leveraged to help with the organization of data, identification of those at highest risk of disease, and targeted interventions based on an analysis of patient populations, then it would make physicians’ jobs easier.
AI might also help solve the doctor shortages by identifying which patients need to see certain in-demand specialists, and who is at the highest risk of complications. Then these patients can be prioritized and treated early, with higher-cost interventions avoided. Preventing heart disease could be a lot cheaper than treating it. Same goes for many different diseases, including cancer.
Lightening The Load
So how can doctors adopt AI? It starts with the clear recognition that there are some things that computers just can’t do. Compassionate care, a sympathetic ear and physical touch are still modalities that work in medical practice. These will never be replaced by a machine. Refining the ability of physicians to effectively communicate these qualities to their patients could improve the overall perception of care, not just for the patient, but also reducing burnout for doctors.
Letting AI do some of the tasks that don’t require a human should lighten the load. Real-time dictation software is already available. But having a computer recognize the essential components of a chart and then doing the data entry automatically would help with one of the major problems in medical care: electronic medical record overload.
Doctors don’t like to have to spend all their time on computer and doing paperwork. Patients don’t like to feel like their doctor is spending their appointment time typing. Letting AI do some of the work can help. It can even suggest diagnoses, and list common treatment plans that are designed to follow the recommended guidelines for each condition.
AI might help to decrease the amount of antibiotic overuse, aid in the recognition of early risk factors for cancer and even match patients with clinical trials for certain medical conditions. The additional information could be used by physicians to assist with the overall care plan.
The world of medicine is bound to change, hopefully for the better. Rather than running from the potential for technology to take over many of the functions doctors do every day, it’s time to fully embrace AI and use its power to keep the humans alive longer than ever before.
About the Author
Kathleen Kozak, M.D., is an internal medicine physician at Straub Clinic and Hospital. She is also a part-time medical director for UHA Health Insurance and is the host of “The Body Show” on Hawaii Public Radio.