Devices that discreetly track our daily activities, exercise minutes, and sleep quality are undergoing a quiet but profound transformation. Once considered exclusively gadgets for fitness enthusiasts, today they are becoming key tools in the early diagnosis and prevention of serious health risks. We live in the era of the medical "Internet of Things," a universe of internet-connected applications, wearable sensors, and smart devices that continuously collect, analyze, and share our health data, opening up unprecedented possibilities for personalized medicine.

According to recent research, nearly a quarter of the population in developed countries uses some type of health tracker. However, wearable devices are just one part of the complex ecosystem that is transforming modern healthcare. Experts like Dr. Sandeep Kishore, an associate professor of medicine at the University of California, San Francisco (UCSF), are leading initiatives aimed at integrating these technologies directly into clinical practice for managing conditions like diabetes and high blood pressure.

A Revolution on Our Wrist: More Than a Step Counter

When we talk about wearable technology, the first association for many is smart wristbands and watches that measure step count, heart rate, and blood oxygen levels. However, the spectrum of these devices is much broader. Smartphones have become central hubs for health data, receiving information from various sensors and enabling direct communication with medical teams through secure platforms. One of the most striking examples is external sensors, such as digital blood pressure monitors or continuous glucose monitors (CGMs).

Continuous glucose monitors, small patches with a nearly imperceptible needle most often applied to the upper arm, represent a true revolution for diabetics. These devices can measure sugar levels in the interstitial fluid every few minutes, sending the data directly to the user's smartphone. In this way, instead of painful finger pricks several times a day, the patient gets a complete overview of glucose trends over 24 hours, allowing for more precise adjustments to therapy and diet. Research teams, like those at UCSF, are actively working on developing systems that will make this data instantly available and actionable for medical teams.

Solving the "White Coat Hypertension" Problem and Other Diagnostic Challenges

One of the biggest challenges in clinical practice is that doctors often get only a momentary, fragmented insight into a patient's condition. Measuring blood pressure in a clinic is a perfect example. It is surprisingly difficult to get an accurate reading in such a setting. The patient might have just had their morning coffee, hurried up the stairs, or is feeling slightly anxious about the visit itself. All these factors can temporarily raise blood pressure, leading to the phenomenon known as "white coat hypertension."

This is where wearable technology shows its true power. Smart blood pressure monitors, which can be used at home, allow for regular measurements in a relaxed environment. They can record blood pressure values throughout the year, not just during semi-annual check-ups. The data is sent to a secure system, providing the doctor with a real and long-term window into the patient's cardiovascular picture. This helps avoid unnecessary therapies and ensures that only those who truly need it are treated. The same applies to detecting atrial fibrillation, where smartwatches can continuously monitor heart rhythm and alert to irregularities that might go unnoticed in a clinic.

The Future is Now: Digital Twins and Smart Diagnostics

The vision for the next five years brings even more advanced concepts. The technology must become simpler and less demanding for the patient. The focus is on solutions that do not require much engagement, such as the possibility of one day measuring blood pressure with a smartphone camera or analyzing heart rhythm via a video recording of a Zoom call.

One of the most intriguing concepts is the creation of "digital twins." This involves an advanced computational model of an individual patient's health. This virtual avatar, built on continuous data from wearable devices, genetic information, and clinical findings, could serve as a testing ground. Medical teams could simulate the effects of different drugs or therapies on the digital twin before applying them to the real patient, predict disease progression, and personalize treatment to an unimaginable degree. Although this technology is still in its early stages and awaits clinical validation, its potential is enormous.

A professor of medicine at UCSF, Dr. Ida Sim, emphasizes the idea of "a bouquet, not the flowers." We will soon face a situation where for a single condition, there will be a range of different gadgets, each collecting its own set of data – "the flowers." This will inevitably lead to information overload. The real challenge and the "secret sauce" to success will be the ability to connect all these different data streams – the flowers – into one meaningful, useful, and simple "bouquet" that will provide doctors with a clear and actionable picture.

Artificial Intelligence as a Key Ally in Medicine

This is precisely where artificial intelligence (AI) comes into play. Dealing with the vast amount of data generated by each patient, which can amount to gigabytes per month, presents a huge processing challenge. AI has the potential to sift through this "avalanche" of data and detect subtle patterns in disease development that the human eye cannot perceive. These patterns can help in understanding the causes of symptoms or even the drivers of the disease itself. The ultimate goal is to turn raw data into clinically relevant alerts and interventions.

The potential of such a system is best illustrated by a real clinical case. Dr. Kishore recalls a patient in her 30s with type 1 diabetes, who needs regular insulin to control her blood sugar. Unfortunately, she ran out of insulin and was admitted to the hospital in a near-comatose state. Her roommate found her unconscious in her room. In the future envisioned by visionaries, such a scenario could be prevented. If the patient had a passive blood sugar monitoring system, the data could be part of a feedback loop between her and her medical team. The system could have sent an alert to a doctor or pharmacist monitoring a dashboard. It could have initiated an automated call or text to her phone, and if there was no response, activated emergency services. Such a tragedy could have been prevented.

Despite fears, AI will not replace doctors. Clinical insight and human experience remain irreplaceable. It is not about a data scientist or an AI expert independently generating clinical insights from a pile of data. Building useful tools requires multidisciplinary teams composed of developers, clinicians, patients, and user experience designers.

Challenges, Standards, and Ethical Issues

The path to this technological future is not without obstacles. Besides the volume of data, aggregation and standardization pose a major problem. Different devices track data in different ways, and companies often use their own proprietary, closed algorithms. This creates "locked silos" of data and makes harmonizing and combining information from different sources a major challenge.

To overcome these problems, projects are being launched such as the collaboration between UCSF and the University of Berkeley to develop an open-source platform called JupyterHealth. The goal of this platform is to merge health data and artificial intelligence for better management of diabetes and hypertension. Using AI models, the platform extracts key insights for doctors and patients in near real-time, enabling decision-making that would take months or years with traditional monitoring.

The key issue remains security, privacy, and ethics. Institutions like UCSF have a rigorous system of checks and balances. A new Health Artificial Intelligence Oversight Committee, composed of experts, reviews projects to ensure that the AI being developed and studied is reliable, fair, safe, and protects people's privacy. Every researcher must submit detailed research plans to an institutional review board, which must approve any research involving human participants, ensuring it is conducted safely and ethically.

Source: University of California