Integrating AI and Wearables into Diabetes Management

Diabetes is a significant and escalating public health challenge. Artificial intelligence (AI) is now being used to enhance early diagnostics and risk assessment for emergent subtypes of diabetes. It is also enabling early identification of at-risk populations and the development of diabetes in unimaginable ways.

Wearable devices, including continuous glucose monitors (CGMs) like the Dexcom G7 and Abbott’s FreeStyle Libre, implantable sensors like Eversense, smartwatches and fitness trackers like Garmin, now integrate seamlessly with CGMs.

In addition, Smart socks for people with diabetes use temperature sensors to detect early signs of inflammation, which can indicate an injury that could lead to a foot ulcer.

These socks continuously monitor key areas of the foot and send alerts to a smartphone App when a temperature change is detected, prompting the user to reduce activity and check their feet. By identifying potential problems early, these socks can help prevent serious complications.

Twin Health uses continuous glucose monitoring and other data from wearable sensors to create a “Whole Body Digital Twin” of a person’s metabolism.

This AI-powered digital replica, combined with a dedicated care team of human health experts in local health centres and district hospitals, can provide personalised guidance to help manage and potentially reverse Type 2 diabetes by managing blood sugar levels, weight, and overall metabolic health.

The system scrutinises how a person’s unique body responds to food, exercise, sleep, and stress to offer actionable insights and reduce the need for medication.

Some patients now use blood sugar Excel Spreadsheets, like the Vertx42, to record blood sugar levels. They then feed Claude or ChatGPT the Excel spread sheet that can include notes on weight, food choices, times of eating, and units of insulin used each day to generate a report and personalised diet plan.

These AI outputs are then shared with physicians at Public Health Centres and State Hospitals via email for feedback and guidance before the next visit to the clinic.

Today, AI-powered systems can analyse vast amounts of data, identify patterns, and make predictions, enabling them to support clinical decision-making and personalise treatment approaches.

This capability to process and comprehend intricate datasets is particularly valuable in diabetes management, where specific patient needs and responses to treatment vary widely.

We now have the option to move beyond simplistic metrics, which cannot capture the rich variability in physiological patterns. Leveraging the full time-series data generated by wearable technologies can provide deeper insights for precise anomaly detection and real-time trend prediction.

The future of this approach hinges on developing scalable algorithms that can integrate contextual data while addressing critical issues such as data privacy and the ethics of AI. Work in these areas will be essential to unlocking the full potential of wearable technologies in diabetes care.

CGMs, smartwatches, and sensors are becoming increasingly crucial in capturing real-time data for individuals living with diabetes. These devices provide continuous monitoring of physiological parameters, allowing individuals to gain insights into their interstitial glucose levels and to make informed decisions about their lifestyle choices.

CGMs have revolutionised diabetes management and provide real-time interstitial glucose readings, enabling individuals to adjust insulin doses, dietary intake, physical activity, and other lifestyle factors to prevent hypoglycemia and hyperglycemia.

AI is changing how people everywhere are managing diabetes by providing personalised treatment plans, real-time monitoring, early warning systems, and virtual health management tools. These innovations help improve glycemic control, prevent complications, and enhance patient self-management and adherence to treatment.

AI algorithms can analyse vast amounts of data, including medical records, lifestyle habits, and genetic information, to create customised recommendations for diet, exercise, and medication. Real-time monitoring and early warning systems, when integrated with wearable devices, can track vital signs such as blood glucose levels, heart rate, and physical activity.

These systems can predict potential hyperglycemic or hypoglycemic events before they occur, sending alerts to patients and healthcare providers to enable timely intervention.

AI-driven decision support systems can help clinicians analyse complex patient data to make more informed decisions about medication adjustments and insulin dosages, and to identify patients at risk of complications.

Hybrid closed-loop systems, sometimes referred to as an “artificial pancreas,” use AI algorithms to automatically adjust insulin delivery based on real-time CGM data, significantly reducing the manual effort required for diabetes management. Deep learning algorithms are being used to analyse medical images, such as retinal scans, to screen for complications like diabetic retinopathy with high accuracy, often faster than manual inspection.

AI models can also predict the risk of other long-term complications, such as kidney disease and cardiovascular events, enabling earlier preventive care. AI-powered chatbots and virtual assistants provide educational content, answer patient queries, set reminders for medication and check-ups, and offer emotional support, making healthcare services more accessible to residents of assisted living facilities.