Digital Health — Connecting the dots
The convergence of technology and medicine has pushed healthcare to the brink of a major disruption. The COVID-19 global pandemic has further brought into focus the central role that healthcare plays in our economic, social and geo-political well-being as a species on Earth.
The digital revolution that has so profoundly transformed the first decades of the 21st century has yet to fundamentally impact medicine. Whereas the market has been the primary arbiter of value driving rapid digital innovation in other spheres, we need to re-imagine digital transformation (Dx) in the entire healthcare value-chain.
There has been much speculation that the convergence of tech and medicine has brought healthcare to the brink of a major disruption which will be the enabler for urgently needed reform. The unsustainability of current healthcare systems has long been recognized, both from cost and inefficiency of delivery, as well as demand due to prevalence of chronic diseases and demographic changes.
In the industrial age, the focus was primarily on managed care in hospitals and expert professional services; In the digital age of medicine, the focus is shifting towards the base with the active engagement of the individual patients and their network (social, family) in governing their health.
What are the drivers for digital in medicine?
Digital drivers of medicine
Pervasiveness of data, big and small. The pervasiveness of digital technologies in modern society has opened access to unprecedented volumes of data, coming from multiple sources. At the same time, data at the individual level, small data capturing personalized physiological parameters, behavior patterns, social and geographical patterns are been recorded from multiple digital sources. This “extended phenotype” made possible by the advent of easily available physiological and behavioral data.
As millions of data points from phenotypes to genotypes become increasingly integrated, it will enable the study of human bodies at multiple dimensions of time and space, from populations to personal level, as shown below.
Pervasiveness of “smart things”. One of the key drivers of digital technologies is the exponential development in computers, sensors, robotics and artificial intelligence (AI), both in power and performance as well as reduction in energy and cost footprints. Most of the existing technologies can be broadly classified as:
Devices: Sensors, wearables and other modalities which are primarily worn or carried by individuals
Lifestyle: Applications, mostly, smartphone or mobile based which allow people to get data and visualization of the various sensing modalities
Social: Social network applications (apps) which try to build a community around specific areas of interests like disease groups, fitness groups
Platforms: Technology enabled platforms which cater to specific sectors of the healthcare industry (like hospitals or medical care provider networks)
The increasing availability of sensors, wearables, smartphones and a plethora of electronic gadgets is opening up new horizons for society. Apart from enabling people to connect in unprecedented ways, these technologies open new avenues to monitor, measure, mine, and model daily activities, lifestyle and habits.
However, the majority of “wearables”, devices and sensing technologies operate in silo, providing “data dumps” to the users rather than “actionable insights”. At the same time, such digitally enable services, particularly for lifestyle and activities data, need to be prioritize security, trust and responsibility, allowing the user transparency and control of the data and its usage.
Advanced Analytics and Machine Intelligence. The value of digital medicine technologies will come from extracting meaning from patient data to generate insights and make decisions which are actionable. The development of advanced techniques for High-dimension analytics (HD Analytics) ranging from causal inference techniques to machine learning and artificial intelligence (AI) methodologies now provide powerful tools to identify patterns in the data and extract actionable insights. An important caveat to take into cognizance is the fact that such techniques are sensitive to the data — thus spurious signals or noise in the data can lead to misleading predictions which can have serious implications in the context of medicine and healthcare.
Prescription for digital medicine
User at Core
Most of the available solutions, either in hardware or software focus on specific tools and functionalities — specific sensing technologies to measure and monitor activities, physiological endpoints or disease states. While these technologies serve as tools to cater to specific needs, they do not cater to a holistic experience for the end-user.
Solutions in the digital health space need to focus on the converge of:
Science & Technology: Leverage state-of-the-art technologies in system science and sensing to provide high precision monitoring.
Behavior & Biology: Solutions should focus on aligning with the behavior, lifestyle, socio-cultural moves of the target user group with a focus on blending in rather than force fit. The ability to provide a natural, fun-filled experience should take precedence over the needs of a specific technology.
Design: Closely linked with the above is the application of design thinking in the process of not only the form factor of the specific devices and user interfaces, but also the overall solution landscape and the specific pain-points addressed by the solution.
Connected Services
The key to creating the holistic experience is connectivity. Digital solutions need to operate in a connected platform where multiple modalities can communicate to provide a desired outcome to the patients, as illustrated above. The large and growing number of technologies, the huge volume of data, and the lack of standardization between multiple software and hardware platforms using different operating systems, programming languages and application environments creates a major friction issue at the level of systems and user experiences.
Digital transformation (Dx) requires an open, flexible and extensible platform which allows integration of information, intelligence and interface across multiple dimensions. Recent studies, c.f. Project Amber, concluded that using single dimension (single biomarker) for complex diseases do not exist (depression and related mental diseases) and concluded, in the context of mental health, “there is value in combining subjective and objective data”. Such studies underscore the importance of flexible and extensive frameworks for Dx in clinical development.
Thus robust and open connectivity platforms are essential to enable integration and allow interoperability between devices, data and analysis to empower users in digital health and medicine.
Connecting the dots
We walk through the user journey of creating a connected experience for multi-dimensional signals (end-points) for users, in this case, patients suffering from diseases of the central nervous system, Alzheimer’s diseases, Parkinson’s disease, dementia and related diseases.
Why?
Diseases of the central nervous system manifest across a broad spectrum. Most traditional therapeutics have little long-term effect in the treatment of such ailments. At the same time, many of the symptoms associated with these diseases, like tremors, mood swings impact the quality of life of patients on a day to day basis.
While specific digital solutions involving wearable devices, or specialized software exist, in many cases, they operate in silos and do not provide a holistic experience for the users or their caregivers.
We embarked on a journey of connecting off-the-shelf devices and services to build a holistic lifestyle monitoring and insights experience for people suffering from diseases of the central nervous system and their caregivers.
What?
Quality of life of people suffering from various diseases of the central nervous system are affected by multiple factors — motor skills, psychological factors, physiological factors (daily activities, exercise routines).
Instead of focusing on a specific device or software, the goal was to develop simple components for each dimension (tremor detection, mood detection, event notification) and thread them together for an end-to-end solution for patients.
How?
Based on Garuda — a connectivity and automation platform, components (called gadgets on Garuda) were developed for specific dimension of signals -
Teramon: A tremor sensing gadget which used off-the-shelf LeapMotion IR sensing device to detect hand gestures and motion which were calibrated for “tremors” typically associated with patients suffering from Parkinson-like diseases.
Game Therapy: A gadget which combines the Teramon gadget with a simple game interface allowing users to interact with the interface rather than the device itself, while the gadget captures and analyzes information from the device to compute different end-points like pinch strength or grab-strength.
Kaomon: Capturing signals from multiple sensory inputs is critical to providing a holistic monitoring experience. This gadget uses another off-the-shelf device, a camera and connected with existing facial emotion analysis system to analyze emotions from pictures taken in the context of a patient.
Events Dashboard: A real-time dashboard to capture the events from various signals (tremors, emotion detection) and connect with alert systems to drive actions
Now, each of these components or gadgets can work individually and provide insights in a single dimension. However, in the context of this user journey, the key component is connectivity.
All the gadgets were connected through the Garuda platform (“garudified”), thereby allowing seamless flow of information between the gadgets — Teramon to Game Therapy, Kaomon to Events Dashboard to drive alerts based on specific threshold for tremor detected by Teramon or based on emotion analysis by Kaomon.
The connected gadgets allow data to be captured in the context of the patients and provide actionable insights for caregivers. The Garuda platform further allows data from other sources (e.g. data available from wearable devices on Daily Steps, Sleep patterns) to be integrated with the existing gadgets for deeper analysis of patients through an Insights Dashboard gadgets — both individually as well as a patient cohort.
See it in action
You can see the workflow in action on the Garuda platform below:
Parting thoughts
It is pertinent to mention that the Garuda platform allows different gadgets to be connected in a plug-n-play, “lego” like fashion. For example, for privacy or security concerns, if a specific scenario prevents the use of the Kaomon gadget, it can be removed from the system without the need to redesign or re- architect the entire flow.
Similarly, if different hardware or software need to be used for tremor detection, the Teramon gadget can be replaced or updated. While different devices or algorithms may not inter-operate individually, the Garuda platform allows them to be connected together for specific workflows. This flexibility is central to the building, deployment and use of digital technologies in medicine, particularly for the touchpoints of patients.
Human life is complex, dynamic and personal. Thus, digital platforms which provide flexibility, configurability and adaptability to changing contexts in healthcare i.e. enable the “connecting of dots”, as illustrated in this user journey on the Garuda platform, are critical for digital transformation.
Footnote: This user journey was developed in collaboration with a major pharmaceutical company in Japan and led to this publication. The demonstration system built on Garuda platform was the seed for the development of a digital therapeutic platform for people with dementia. More on that story in another post!