4 Ways Artificial Intelligence is Transforming Healthcare

Faculty Insights Industry Advice Computing and IT Healthcare

Julian He, Senior Data Scientist at Insulet Corporation, discusses four ways that artificial intelligence is impacting the healthcare industry.

4 Ways Artificial Intelligence is Transforming Healthcare

Artificial intelligence (AI) is changing the healthcare landscape. It has become one of the major driving forces in patient care innovation, especially in chronic disease management for conditions such as diabetes. Currently, AI is accelerating four major trends in healthcare innovation:

1. Mobile Platform Adoption

Mobile apps are used by many patients and care providers. As such, optimized mobile platforms can help drive patient engagement, facilitate more efficient patient data collection, provide patients and caregivers easy access to data, and improve communication between patients and patient support groups (relatives, nurses, etc.).

2. Continuous Patient Monitoring

In 2018, clinical trials by the University of Chicago Medical Center proved that continuous glucose monitoring (CGM) is cost-effective for Type 1 diabetes, which affects about 10 percent of the total population living with diabetes. CGM helps patients to improve overall blood glucose control and reduce diabetes complications. Without CGM, patients depend on the manual measurement of blood glucose levels several times a day. This kind of data cannot provide a full picture of a patient’s status and is based only on a selection of measured time. Patients with a CGM device, however, can monitor their blood sugar level once every five minutes. This is only one example of how continuous patient monitoring is being used to save money, improve health outcomes and increase the overall quality of patient care.

3. Intelligent Diabetes Management

Diabetes management is very complicated due to the variation between each patient’s insulin sensitivity, lifestyle, activity levels, and diet habits. It is hard to get a one-size-fits-all solution in the early development of treatment plans. Artificial intelligence now allows clinicians to use patient data to develop personalized treatment recommendations to help manage the disease. Many intelligent systems target a specific task during a diabetes management procedure, such as the prediction of blood glucose levels, prediction of hazardous complication events, prediction of patient re-admission, etc. These “intelligent” treatment plans that use patient data are likely to have better success rates, as they can simplify the complexity of diabetes into a manageable problem statement.

4. Automatic Insulin Delivery

The optimal goal of diabetes management is to have a system that can automatically deliver insulin when it is needed, just like a pancreas in a healthy person. Automatic insulin delivery is similar to autonomous, or self-driving, car. This technology is still undergoing clinical trials and is under review by the FDA; however, a hybrid version (one that still requires human intervention), the Medtronic MiniMed 670G Hybrid Closed Loop, was approved by FDA in 2018.

In summary, the advancement of artificial intelligence in diabetes management has been accelerated in recent years. Many traditional companies have invested heavily in the new innovative technology according to their product pipelines.  Some up-and-coming startups, such as Glooko, Tidepool, and Bigfoot biomedical, have brought disruptive ideas to the market place and may be able to reshape the landscape of diabetes care.

Overall, a breakthrough in artificial intelligence in healthcare is coming. Engineers in artificial intelligence are in high demand. Are you ready for the trend?