3 Ways Artificial Intelligence Is Transforming Healthcare (Part 4)

If you’ve been keeping up with our “AI Transforming Healthcare” series, you’ve learned about some cool ways that AI is impacting the medical field; selfies as diagnostic tools, electronic health records as risk predictors, and better brain-computer interfacing are just a few examples we covered. In case you missed our previous entry, you can get up to speed here.

In this fourth and final part of the series, we’ll go over how AI is improving cancer treatment, enhancing access to resources for patients in rural and underserved areas, and creating better end-of-life tools for elderly and ill patients.

Advanced Immunotherapy for Cancer Treatment

Cancer attacks the immune system and stops it from attacking tumor cells. As a result, immunotherapy is one of the best cancer treatment methods right now. It strengthens the patient’s immune system and helps stop tumors from spreading. Unfortunately, most patients don’t respond to immunotherapy, and oncologists don’t exactly know why.

An AI application could potentially help elucidate this quandary; oncologists could identify patterns in patients who do respond well to immunotherapy and figure out if it’s a viable treatment option for others. But this all depends on the data.

According to Dr. Long Le, MD at the MGH Center for Integrated Diagnostics in Boston, the key to success is massive amounts of patient data. Any patient undergoing new therapies should opt into sharing their data so that other patients can benefit. Sharing patient data between hospitals, states, and even countries could further accelerate therapy results.

And the field is still growing with knowledge and new experimental treatment options, making things much more complicated. “Recently, the most exciting development has been checkpoint inhibitors, which block some of the proteins made by some types of immune cells. But we still don’t understand all of the disease biology. This is a very complex problem,” he says.

Addressing a Lack of Resources

During a 2018 World Medical Innovation Forum (WMIF) panel, speakers pointed out that more radiologists work in Boston’s hospitals than in all of West Africa. And this disparate spread of doctors isn’t the only problem. We’re facing a worldwide shortage of doctors, technicians, and radiologists. Lack of information and knowledge kills more people than disease alone.

With AI, some diagnostic tools could be replaced to give the current batch of doctors time to see more patients. Jayashree Kalpathy-Cramer, Ph.D., thinks this type of technology has immense potential to increase access to healthcare.

But, as Dr. Ziad Obermeyer, an Assistant Professor of Emergency Medicine at Brigham and Women’s Hospital, also touched upon, these algorithms must not contain any bias or single-population training data. They must, if possible, include data from patients all over the world and in every socioeconomic class.

Kalpathy-Cramer says that disease and populations can vary wildly from country to country, which underscores the importance of an unbiased, fully-trained algorithm. “As we’re developing these algorithms, it’s very important to make sure that the data represents a diversity of disease presentations and populations – we can’t just develop an algorithm based on a single population and expect it to work as well on others.”

AI’s Bedside Manner

Healthcare’s shift to focusing on the baby boomers entering old age will bring with it a new set of challenges and requirements. Predicting disease, risk factors, and worsening symptoms are the main priorities for any doctor, but AI can help speed up processes.

Using predictive analytics and decision-making tools, providers can be alerted to patient issues well before the patient comes in for a visit. Not all illnesses are able to be monitored 24/7, but more common conditions, like sepsis or seizures, can be tracked with an algorithm that’s fed tons of complex data.

AI will help support decision-making about continuing care for patients whose condition has deteriorated or are critically ill. For example, patients in a coma after cardiac arrest are unable to speak for themselves. In situations where family members aren’t available or alive anymore, AI can help determine whether the patient could pull out of the coma or if they should go peacefully.

Dr. Brandon Westover, MD, says that doctors are required to check over the EEG data visually. But it takes a lot of time, and the accuracy rate varies between each provider; as such, patients may receive differing opinions from various doctors, which can be incredibly stressful.

“In these patients, trends might be slowly evolving. Sometimes when we’re looking to see if someone is recovering, we take the data from ten seconds of monitoring at a time. But trying to see if it changed from ten seconds of data taken 24 hours ago is like trying to look if your hair is growing longer,” he says.

However, AI can shed new light on patients’ conditions. Using an algorithm and varied training data across many patient populations, AI can find patterns and generate actionable insights for providers. Subtle signs of improvement could go unrecognized by a tired doctor, while an AI algorithm could detect it immediately, giving the patient a better chance of recovery.

Smarter Healthcare Is Coming to a Hospital Near You

We hope you’ve enjoyed this series on how AI is transforming healthcare. This technology is already having a profound effect on health tech development. But this is just the beginning. With opportunities to improve cancer treatment options, access to resources, and monitoring of patients, the possibilities with AI are really endless.

Which AI application discussed in this series has the most potential to advance medicine? What other areas of healthcare are in dire need of some AI innovation? As always, let us know your thoughts in the comments below!