Modern technological innovations are continuing to suggest promising approaches to the digital operations of contemporary health systems. These approaches are enabling the next generation of learning health systems, with technologies such as real-time data streaming, clinical decision support, and robust artificial intelligence powering novel and innovative supporting improvements to health system care. However, the implementation of these modern systems remains challenging, often requiring complex restructuring of technical capabilities and capacities. The HeartAI team propose HeartAI as a modern approach to these contemporary challenges, promising best practice implementations and insights to support the modern digital health system.
Supporting digital health
HeartAI hopes to support health system care by increasing the capabilities and capacities of digital health. HeartAI ecosystem development achieves this by providing modern approaches to resource management, platform orchestration, security implementations, data integration and access, reporting and visualisation, powerful and real-time analytics, and many opportunities to support clinical service delivery and health system management.
HeartAI is primarily a cloud-native reactive microservices architecture that is orchestrated with the Red Hat OpenShift platform and the actor-based concurrency system Akka. These capabilities are enhanced with best-practice deployments to Microsoft Azure, extensible software-defined networking, modern identity and access management, real-time logging, monitoring, and observability, modern service middleware, high-performance analytics, and robust deployment frameworks.
HeartAI implements a variety of contemporary software solutions. This includes software to support cloud environment management, container orchestration platforms, logging and monitoring capabilities, identity and access providers, and software-defined networking, among others. The following documentation provides an overview and reference to these technologies within HeartAI ecosystems generally:
Application software is integrated with health information systems and HeartAI data services, allowing rigorous and real-time data and information to be readily accessible. This enables targeted insights to be generated from these systems, with operational capabilities and practices to support rapid response to health system needs. HeartAI deploys real-time capable visualisation software with support for data streaming and event-driven behaviour, allowing continuous engagement with clinically important information and activity.
HeartAI services are core middleware implementations that support general digital health functionalities. This includes service capabilities such as data services, linkage services, aggregation and reporting services, and analytical services. The implementation of these services considers reactive microservices architectures and follow concepts from The Reactive Manifesto. Service architectures benefit from modern design concepts such as reactive design patterns and event-driven architectures.
Services often represent corresponding domain models, for example the HIB interface service implements domain functionality for the HeartAI interface with the SA Health Health Information Broker (HIB). HeartAI service implementations provide high-performance extensions to the digital ecosystem, supporting comprehensive data integration, data-streaming, and high-availability.
Evolving approaches with probabilistic programming and computational statistics are promising powerful methods to define and create analytical systems. These methods can represent probabilistic models under a variety of conditions and constraints, and are capable of simulating new data from these states by the generative nature of these models. HeartAI deployments supports inference from a Bayesian perspective, such as by generating samples from a prior or posterior predictive distribution, where there may also be a conditioning or marginalisation of such a distribution. In addition to the creation of a representative probabilistic construct, these models may be used for the generation of new data under a variety of assumptions and hypothetical situations, and allow for the prediction and forecasting of future events and potential outcomes.
Supporting the SA Virtual Care Service, HeartAI has developed a variety of probabilistic approaches that model the service activity generally and specifically to support personalised patient care. This includes the creation of models for admissions to the service over time, including the patient population of admissions, the likelihood of potential outcomes for the individual patient, and the length-of-service for each episode of care.
HeartAI proposes to support the South Australian digital health system with an evolving ecosystem implementation and deployment, with functionalities for real-time data integration and processing, adaptive and intelligent data linkage, dynamic and interactive data reporting, powerful modern analytics, and user-interfacing applications and dashboards. The HeartAI team are progressing with the following current exemplar project initiatives:
HeartAI developers are supported with development environments that encourage a positive and productive development experience. HeartAI instances are deployable to a developer’s local machine with automated installation steps, and these instances are secure, lightweight, and reproducible. The HeartAI development environment is designed to be modular, with developers deploying corresponding components to support their development objectives and use-cases. This typically includes integrated development framework components, such as Git, Git LFS, Docker Engine, and Docker Compose, such that the developer may interact with the HeartAI source respository. These integrated deployment components also support pre-declared software module packaging and deployment and a variety of backing services including logging, monitoring, networking, database, and identity services.
In addition to local machine installations with a wide-variety of configuration options, HeartAI instances also deploy to Red Hat OpenShift with particular support for managed Microsoft Azure Red Hat OpenShift environments.
The following policy documents describe and specify the responsibilities and obligations of HeartAI environments and HeartAI administrators and developers within the health system context. HeartAI is expected to maintain a high-level of due diligence, particularly with the management of potentially sensitive information. HeartAI is committed to close coordination with SA Government and SA Health organisations and to understanding and following state and federal laws, policies, regulations, and compliance standards.
The HeartAI team have matured positive and productive relationships with key stakeholders locally within South Australia, nationally, and internationally. These engagements support the development of an ecosystem that hopes to bring together digital health, academia, industry, and the greater community.
Many friends and colleagues have supported the development of HeartAI. The HeartAI team are thankful and grateful for these relationships.