November 14, 2023 - Julie Meslin, MS, OTR/L
November marks the annual National Awareness Month for both chronic obstructive pulmonary disease (COPD) and diabetes. According to the Centers for Disease Control and Prevention, 15.7 million Americans are diagnosed with COPD and 37.3 million Americans are diabetic. The effect upon the daily lives of those with these diseases, as well as their caregivers and loved ones, cannot be overstated.
One way the treatment landscape is evolving to improve outcomes is through the realm of artificial intelligence (AI). Here, we focus on a few ways that AI plays a role in detection and management of these conditions.
Remote Monitoring
In the case of diabetes management, the first CGM (continuous glucose monitoring) systems were available to the public in 2000. These monitoring systems have evolved into “closed loop systems” or what is referred to as an “artificial pancreas.” The CGM, worn by the user, monitors blood glucose levels in real time and communicates this information to AI control algorithm software, which processes the raw data and predicts insulin dosage levels within a target range.
The wearable insulin pump delivers both basal (background) and bolus (mealtime) doses. Patients can download apps to their smartphones that provide them with information about glucose levels, including real-time data and historical trends. Apps provide a user-friendly way for patients or caregivers to access information, allowing them to be proactive about self-care.1
There are also wearable AI devices for those who live with COPD. Wearable sensors that communicate with apps can monitor activity, sleep, respiratory rates, and oxygen levels. As is the case with CGMs, the data collected through these devices allow the user to access real-time data as well as share this information with family, caregivers, and health care providers. The availability of data can reduce the need for costly and inconvenient in-person visits.
Detection and Predictive Analytics
AI also aids in the early detection of diabetes and COPD. In a recent study, an app-based spirometer was deemed “a simple and adequate tool for early COPD detection in outpatient clinics.”2 Another way AI is used in early detection is by analyzing vast amounts of data quickly and accurately. Al algorithms process medical records, test results, family history, and lifestyle factors to identify those who are at a greater risk of developing a chronic disease. This empowers health care providers to identify patients early or perhaps even prevent the development of the disease altogether. Predictive analytics can be used to examine historical patient data to predict risk of hypoglycemic or hyperglycemic events, providing advanced warning so that the patient takes proactive measures. For example, data collected by activity trackers, merged with data from CGMs, smart pumps, and other sensors, can be used to automatically detect exercise sessions and make dietary recommendations to avoid hypoglycemia.3
Daily Management and Lifestyle Recommendations
The complicated medication schedules that are part of the daily lives of people with COPD or diabetes are mitigated by using AI in the form of real-time data monitoring and reminders. This eases the burden for patients and caregivers. AI-powered apps can offer exercise plans, stress management techniques, and dietary advice, as well as reminders and incentive structures such as gamification, points, and badges that reward the patient for adherence, which increases motivation.
Considerations in AI Technology for Chronic Conditions
Although AI is a boon to the treatment of chronic conditions like diabetes and COPD, there are potential risks that must be considered. “Despite tremendous strides made in the field of AI…it is not accessible to all societies,” authors Farhud and Zokaei explain.4 “Many low-income and developing countries still do not have access to the latest technologies. It should be noted that the ethical dilemmas, privacy and data protection, informed consent, social gaps, medical consultation, empathy, and sympathy are various challenges we face in using AI.”4
Health care providers must not over-rely on these powerful technological tools. Providers’ human connection to their patients remains the cornerstone of care for building trust, fostering open communication, and informing patient-centric decisions in treatment planning. Incorporating AI into diagnosis, treatment, and daily management has the capacity to inform holistic care that is efficient, flexible, and grounded in evidence.
References
- Kebede MM, Pischke CR. Popular diabetes apps and the impact of diabetes app use on self-care behaviour: a survey among the digital community of persons with diabetes on social media. Front Endocrinol. 2019;10:135. doi:10.3389/fendo.2019.00135
- Lin C-H, Cheng S-L, Wang H-C, et al. Novel app-based portable spirometer for the early detection of COPD. Diagnostics. 2021;11(5):785. doi:10.3390/diagnostics11050785
- Vettoretti M, Cappon G, Facchinetti A, Sparacino G. Advanced diabetes management using artificial intelligence and continuous glucose monitoring sensors. Sensors. 2020;20(14):3870. doi:10.3390/s20143870
- Farhud DD, Zokaei S. Ethical issues of artificial intelligence in medicine and healthcare. Iran J Public Health. 2021;50(11):i-v. doi:18502/ijph.v50i11.7600