Although enterprise-wide AI will begin in the traditional IT department, few departments, however, are prepared to take on the complexity and challenges that becoming AI-enabled will pose. Much like the rest of the organization, AI will require upskilling, modernizing legacy architecture, managing data more tightly, and establishing new practices.
Health care is, if nothing else, built upon outdated hardware and software. This legacy architecture will introduce significant risk to AI efforts when older systems are not robust or secure enough to support data-intensive AI capabilities. Modernizing the architecture will require upgrading or replacing some existing technologies or deploying new hardware and applications, whether on-premises or in a hosted cloud environment.
To have seamless AI workflows, modern GPUs for highly parallel workloads (rather than CPUs,) faster networks (>100 Gbs,) and fast storage for highly-scalable, low-latency storage optimized for parallel access are needed. Modern data hubs that break down the traditional silos to allow for easy movement of the data is absolutely required.
According to IDC research, 60% of participating health care organizations lack a data strategy. A Gartner survey found that concerns about data quality and scope are the third ranked challenge to AI adoption, behind lack of talent and not understanding use cases. It is reasonable to suggest that such high-performance computing work has been and continues to be beyond the core competencies of either healthcare organizations or governments.
Health data infrastructure, defined as the hardware and software to securely aggregate, store, process and transmit healthcare data. Currently, health systems have fragmented data within their organizations and the ability to transmit this data to outside entities is severely limited. This situation evolved as individual organizations had to buy and maintain the costly hardware and software required for healthcare, and has been reinforced by vendor lock-in, most notably in electronic medical records (EMRs). With increasing cost pressure and policy imperatives to manage patients across and between care episodes, the need to aggregate data across and between departments within a healthcare organization and across disparate organizations has become apparent not only to realize the promise of AI but also to improve the efficiency of existing data intensive tasks such as any population health and patient management.
To fully take advantage of everything AI can offer in healthcare, it is imperative to use aggregated healthcare data to produce powerful models that can provide their outputs within the clinical workflow of the clinicians in a timely manner, such as automated diagnosis and also enable an increasingly precision approach to medicine by tailoring treatments and targeting resources with maximum effectiveness in a timely and dynamic manner.