Industry
Healthcare
Patient-centered Care: Health and Wellness Redefined
Services
IoT and Digital Engineering
Bringing Together IoT and Digital Engineering to Create Rich Experiences
Data and Analytics
Harness the Power of Data and Analytics for better Business Outcomes
Highlights
- Alzheimer's is the most expensive disease in America, and nearly USD 600 billion per year is spent in direct costs in US alone.
- This paper discusses a platform approach—possible non-pharmacological interventions that can be managed using digital health technologies—reduce the disease burden and positively impact patients.
- A mash-up of sensor data, electronic patient-reported outcomes (ePROs), clinical data, and physician’s inputs help create a holistic end-to-end care model, empowering patients with Alzheimer's disease.
- Future opportunities in areas like AI-based personalized treatments, holistic diagnosis, native wearables, and drug discovery, could play a critical role in minimizing the impact on independent living.
In this article
In this articleinpage
At a glance
Alzheimer’s disease (AD) has one of the highest economic burdens
It costs nearly USD 612 billion per year—USD 355 billion in direct costs and USD 257 billion in indirect costs in the US alone. Over 16 million people provide unpaid care for family members affected by AD and other types of dementia. The costs are further increasing in low-income nations. Hence, reducing this cost burden is an important problem to solve.
This paper discusses a platform approach, one that leverages digital health technologies, to manage the burden of AD.
The evolving market of digital health technologies
As per a market study, the therapeutic segment for the global digital biomarkers for neurodegenerative diseases is 20.63%
By 2025, payers are expected to emerge as the largest end-user segment in the global digital biomarkers market. The growth rate for this segment is expected to be 44.04%, during the forecast period―2019 to 2025. This demonstrates a large market potential for digital health technologies in the segment.
Possible non-pharmacological interventions can be managed well using digital technologies, which positively impact the patients, thereby reducing disease burden.
Managing AD with digital biomarkers
Digital biomarkers are verifiable data on disease manifestations that are measured using digital technologies
These can be used for two key purposes:
- Early detection of AD, dementia, or precursors like mild cognitive impairment (MCI). This can be done by leveraging longitudinal analysis—the practice of analyzing human or patient data over time—within Internet of Things (IoT) environment. However, this mostly works on a captive population. Studies in the wild did not produce comprehensive results. Though, the work can be extended using time-series analysis methods like Markov chains and the like.
- Management and support of Alzheimer’s patients using sensors and artificial intelligence (AI), to improve quality of life.
We envisage that both these applications will coexist. The first use case will enable effective screening and the second one will enable effective management. For example, a navigation mobile application can ascertain if a patient has lost his or her way by tracking their movement patterns and guiding them home. Another application can be a face recognition system that maps the face to a name or a relationship. This can be done using a smart glass that understands when the patient cannot recognize a known face.
A platform approach to effectively manage AD
There are challenges in the adoption of digital biomarkers, in the current clinical practice
However, recent works have shown promise in the management and therapy of chronic diseases, using digital health.
A platform approach to the problem entails a mash-up of sensor data, electronic patient-reported outcomes (ePROs), clinical data, and physician’s inputs to create a holistic end-to-end care model. This will enable AD patients to deal with severe challenges in their activities of daily living (ADL) within their environments.
Key considerations in designing such a platform include:
Barring the demographic differences between patient populations, inter-patient variation exists in both the symptoms (of AD and dementia); it is also crucial to factor in the infrastructure and devices that are available to the patient.
Real-time requirements exist for some interventions like understanding the ‘walk versus wander’ patterns. This makes it necessary to use on-device computing and inferencing techniques to reduce latency.
These patients are likely to forget to charge their wearable devices or put them on after charging. Hence, a mechanism of battery-less wearable and object sensorization can be valuable for AD patient monitoring.
Multi-device and multi-tenant scenarios demand an ecosystem approach to the problem, since device vendors, physicians, providers, and patients are all stakeholders.
Figure 1: A platform approach to managing AD
Table 1 is a concept map that defines the functional features and provides aframework to design a complete use case for a feature (see Figure 1).
Table 1: A concept map
Proposed architecture for an AD interventional platform
The platform access layers will enable consumers to avail services securely
This includes web interface, apps, application programming interface (API), and other workflow management services like consent management, alert management, and reporting modules.
The service desk offers an integrated facility with alternative means of providing services to the patient and kin, especially for the non-tech-savvy, and where manual intervention is required.
To build stakeholders’ confidence, security, audit trail, and regulatory layers are embedded, ensuring basic hygiene. These layers extend to realize software-as-a-medical-device (SaMD) requirements.
The interoperable IoT integration layer will enable seamless device (vendor-neutral) integration and can perform real-time or batch complex event processing (CEP). The layer is a state-of-the-art IoT framework that can handle time-series sensor data, and multimedia data, and host multiple AI models. These models can be initiated on-demand to enable intervention or improve effectiveness, as depicted in table 1.
The platform provides a holistic service offering to consumers and connects them to providers, making services ubiquitous.
Figure 2: Platform access layers
Figure 2 depicts the critical elements of the platform that will enable the defined interventions for an AD patient
The platform access layers will enable consumers to avail services securely. This includes web interface, apps, application programming interface (API), and other workflow management services like consent management, alert management, and reporting modules.
The service desk offers an integrated facility with alternative means of providing services to the patient and kin, especially for the non-tech-savvy, and where manual intervention is required.
To build stakeholders’ confidence, security, audit trail, and regulatory layers are embedded, ensuring basic hygiene. These layers extend to realize software-as-a-medical-device (SaMD) requirements.
The interoperable IoT integration layer will enable seamless device (vendor-neutral) integration and can perform real-time or batch complex event processing (CEP). The layer is a state-of-the-art IoT framework that can handle time-series sensor data, and multimedia data, and host multiple AI models. These models can be initiated on-demand to enable intervention or improve effectiveness, as depicted in table 1.
The platform provides a holistic service offering to consumers and connects them to providers, making services ubiquitous.
Addressing platform adoption challenges: Key factors
As healthcare is digitalized, building an ecosystem of service providers, including clinicians, hospitals, and specialists is vital
The digital continuum of care ensures instant response to an event. Actionable insights sent to various stakeholders and faster response to alerts will improve confidence in the system. Data interoperability will be crucial in developing a holistic and secure platform for AD patients with continued focus on patient experience.
Usability, performance, and availability will help in creating an exceptional experience. Beyond all the right privacy features, security provisions and alignment to the regulatory policies and procedures will ensure better adoption.
The platform architecture illustrated in this paper (Figure 2) could be extended to various NDDs (neuro degenerative disorders), enabling patients to avail required services instantly.
Future opportunities lie in areas like AI-based personalized treatments, holistic diagnosis, native wearables, and drug discovery. AI models can combine sensor data, patient, and caregiver-generated text or speech data, and categorical data generated from lab findings with magnetic resonance imaging (MRI), or computerized tomography (CT) data from imaging to create a comprehensive and holistic patient model. Real-time actionable insights extrapolated by leveraging edge and on-device computing could play a critical role in minimizing the impact on independent living.
There is also a clear opportunity to build native wearables and custom devices in wearable or implantable forms using hardware-software co-design. Furthermore, developing a digital twin of the human brain could potentially enable the creation of simulation models to understand the kinetics and biochemistry of the disease, leading to more effective drug discovery. We can safely conclude that the future looks promising as digital health technologies present an enormous potential to reduce the economic burden of AD and redefine chronic condition management.
