Joshua R. Windmiller
Joshua Windmiller is a seasoned technology innovator and entrepreneur specializing in medical devices and diagnostics. As the Head of New Analytes and Sensing Platforms at Dexcom, he leads a team of cross-functional experts shaping the future of wearable sensing technologies. Previously, he co-founded Biolinq, where he served as Chief Innovation Officer, Chief Technology Officer, and a Board Member, driving advancements in biosensing technologies. With a strong background in interdisciplinary engineering, he excels in translating innovation into successful businesses and large-scale projects.
Through this article, Windmiller highlights how advancements in semiconductors have been instrumental in the evolution of medical wearables, enabling increased connectivity, sensor integration, and AI-driven predictive analytics to transform raw data into actionable health insights.
"Future frontiers in medical wearables – Where are we going, and what can we learn from the semiconductor industry?
The aphorism “it has never been a more exciting time to be in the field” appears to be common parlance in many technology disciplines. While this statement is used so frequently that it qualifies as cliché, it could not be more applicable to the breakneck evolution of modern medical wearables.
Over the past three decades, medical wearables have benefited from advancements in many scientific and technical fields. Among these, no field owes greater credit to the widespread proliferation of medical wearables than semiconductors. Indeed, developments in highly complex ICs have served as foundational enablers for medical wearables to help individuals lead healthier, happier, and more productive lives, lessening the burden of the management of chronic disease, facilitating more rapid or decentralized diagnosis of a physiological state, and empowering individuals to achieve their personal wellness goals. As an enabling tech stack, semiconductors have imbued medical wearables with the wireless connectivity required to keep the wearer’s healthcare provider and support network informed. Likewise, semiconductors have enabled miniaturization to incorporate advanced sensory functions onto a body-adorned, power-constrained wearable. Innovations within the embedded processing domain have recently provided the computing power requisite for advanced signal processing algorithms and AI/ML functions directly within the wearable. Indeed, medical wearables would be impossible without precision analog front ends, mixed-signal wireless radios, and power-efficient yet high-performance microcontrollers supporting multiple peripherals and functions.
Past and current state of the industry:
The medical wearables of yesteryear were often mere data loggers, aiming to aggregate measurements to be later reviewed by the healthcare provider or to render an alert upon detecting a critical event. While certainly lifesaving under many circumstances, these devices represented a true Phase 0 of the field.
Phase I – The drive to increase connectivity
Incorporating low-power wireless radio technology (e.g., Bluetooth Low Energy, Wi-Fi, Near-field Communication, cellular) represented the first leap in the utility of medical wearables. Imbued with wireless connectivity, wearables were – and continue to be – able to provide users with real-time data or targeted alerts to instigate proactive measures to manage (or preserve) health. Cloud-based connectivity has also facilitated compliance tracking to prescribed therapies or provide insight into patient behaviors to ensure the appropriate healthcare interventions are taken.
“The next evolution in the field has been playing out over the past few years. Sensor fusion—augmenting traditional capabilities with multiple dimensions of electrophysiological, pathophysiological, kinesthetic, and electrochemical time-series data— represents an emergent thrust within the industry.”
PHASE II – Sensor augmentation & integration
The next evolution in the field has been playing out over the past few years. Sensor fusion—augmenting traditional capabilities with multiple dimensions of electrophysiological, pathophysiological, kinesthetic, and electrochemical time-series data—represents an emergent thrust within the industry. Incorporating these various sensing modalities necessitates the advent of advanced chip-scale integration techniques capable of combining multiple semiconductor technologies monolithically or otherwise requires advanced packaging methods to combine these disparate technologies into a compact, PCB-mountable assembly or module.
Industry trends going forward:
Phase III – Moving from data to wisdom
Those familiar with the DIKW hierarchy will appreciate that data, although the basis for all judgment, is oftentimes a weak instigator of action. As data is organized into information, users can begin to understand trends and piece together a story of events that transpired, much like reading a periodical. Over time or multiple channels, the synthesis of information can be used to build knowledge, empowering the user to link action with outcome. Finally, we have wisdom, enabling us to think ahead and predict outcomes before actions are taken. Viewed from this lens, translating data into targeted, actionable insights that are contextually relevant for the user will comprise the next medical wearables evolution phase. Using predictive analytics and statistical learning techniques, complex time series and multi-dimensional datasets can be condensed into key learnings to drive more immediate medical interventions, improve outcomes, or encourage healthy behaviors.
When examined from the perspective of wearables, in some scenarios, migrating from raw data to targeted insights cannot be borne by the Cloud due to latency, patient privacy, and cybersecurity concerns. More sophisticated ultra-lowpower embedded processors capable of performing these highly parallelized inference generation routines are needed and represent a new, untapped synergy between semiconductors and medical wearables.
What can we learn?
This retrospective (and prospective) provides a few key themes regarding using semiconductors in medical wearables. Firstly, we will continue to see continued progression in implementing wireless radios, offering increased data reliability/availability, lower power operation, increased range, and more secure encrypted communication channels. Secondly, medical wearables will witness the continued expansion of their sensory capabilities, ingesting additional real-time data streams, hence emphasizing the importance of heterogeneous integration and novel packaging methods. Lastly, exciting developments in AI/ML are increasing demands on local, ultralow power computing, providing users with a delectable wear experience that empowers them to manage their health best.
In summary, with all the advancements in enabling domains, perhaps the statement, “It has never been a more exciting time to be in the field,” can be perceived as a truism. The future certainly looks bright for medical wearables.
The views and opinions expressed in this article are those of the author and do not necessarily reflect the positions of the author’s employer, Dexcom, Inc.
