AI early-stage companies are developing applications to help clinicians and staff with a range of work activities, from automated administrative work that reduces documentation time and frees up clinician hours, to predictive analytics and real time computer vision to improve OR efficiency. To fully leverage the value of these exciting developments they will need endpoint hardware capable of delivering a reliable and secure workspace.
Healthcare staff and clinicians travel between various locations in a healthcare system and between floors within a hospital. They already have experienced the frustrations of being unable to access data at a critical care moment due to a system being down or an incompatibility issue between desktops. Consistent performance at any care location is essential.
Whether AI-driven or legacy applications, the endpoint is the decisive moment. As AI increases in use, having hardware and software that supports a high-performance experience for all staff will improve operational efficiency, patient outcomes, and revenue.
Key issues to consider are hardware choices, AI readiness, security and compliance, operational interoperability, and total cost of ownership (TCO).
Hardware options in the AI dnvironment
Within the mix of endpoint devices healthcare systems use, PCs and thin clients are two common choices. As systems continue to face tight budgets, thin clients are an economical option, requiring less IT staff time and costs than PCs. Hospitals working in a variety of computing environments can use thin clients to support VDI, cloud, hybrid and on premises environments, thereby extending the ROI of their investment.
Housed in a small form factor, an advantage in busy hospital and clinic settings, modern thin clients are fanless and solid state to promote reliability, durability and energy efficiency. They require low maintenance and can be centrally managed, cutting down on administrative time.
The advent of AI is adding AI PCs to this mix. These can require neural processing capability and function at the higher end of CPU/GPU compute power, further driving up energy costs. As a result, they come with high price tags that can be prohibitive for budget-conscious healthcare systems. Traditional PCs tend to fall in the mid-range in cost but usually require a refresh every three to five years, a frequent drain on budgets.
Thin clients are read-only devices and do not store any local data. When evaluating thin client options consider what a staff or clinician’s needs are to determine the best choice in processing power, screen quality, number of ports and other elements. Also, interoperability with major VDI vendors like Citrix, Omnissa and Microsoft is a must.
A longer refresh cycle is a thin client benefit, helping to contain costs and protect TCO. However, thin client vendors vary on what is covered in a standard agreement. Bolt on features that will run up costs include healthcare specific subscriptions based on enterprise models, upgrade charges and support subscriptions through an endpoint lifecycle.
These substantial extra charges may reach the point where rip-and-replace looks like a better option, requiring new purchases and detracting from sustainability goals by adding to waste.
As AI plays a deeper role in how clinicians and staff serve their patients, healthcare systems are examining the best hardware devices to support AI workflows. The volume of AI data in large language models is processed mainly in the cloud or on servers. Locally run AI applications are still limited by the cost of AI PCs. Thin clients, built to support VDI and the cloud, can provide a low power, lower cost gateway to access AI workload resources. Traditional PCs still offload AI workloads to the cloud and come at a higher cost.
Securing the endpoint
Healthcare system executives face the triple threat of criminals now using AI to execute a breach, PHI data leeching into AI large language models and severe financial losses due to successful cyber-attacks. Add to this HIPAA violations and the losses worsen.
Cyber criminals favor healthcare which has seen a steady rise in attacks: from 2019 to 2024 healthcare data breaches rose annually from 397 to 536, exceeded only by the financial sector. Healthcare breaches still took the longest to identify and contain, averaging 279 days, according to DemandSage. It is a main reason, HIPAA Journal notes, for the sector’s popularity, and healthcare data being so highly valuable. The stolen data can be used for longer than a stolen credit card which can be cancelled right away.
Countering new threats requires an increased emphasis on guard rails to restrict what data can populate large language models. At the endpoint, locked down, read-only thin clients help reduce the attack surface since they do not store data locally. Another option, zero clients, have no local OS, and connect to remote desktops. These devices mitigate risk as busy clinicians and staff move between locations, remotely accessing medical records and other PHI data.
To enhance data security, healthcare IT will also want to integrate thin clients (or PCs) with software that provides another protection layer. Interoperability with Imprivata Enterprise Access Management with Single-Sign-On will enable advanced identity authentication and support HIPAA compliant reporting. For clinicians it provides passwordless authentication so they can authenticate once, then re-access applications quickly as they move between workstations.
A secure AI future
Over the next few years AI powered computing will become the foundation of healthcare systems, and much of it will begin at the endpoint. No doubt AI PCs will eventually become commoditized as did traditional PCs. However, the compute and power costs will continue as well as more frequent refresh cycles. An alternative with lower TCO and less attack surface will continue to be thin and zero clients. Healthcare executives envisioning their AI future will look at a mix of thin clients and AI PCs to provide superior patient care and sustainable business outcomes.
Photo: Weiquan Lin, Getty Images
A computer science graduate with numerous IT certifications, Kevin has more than 25 years of experience in the IT sector, including remote connectivity, terminal emulation, VoIP, unified communications, and VDI remoting protocols. Since joining 10ZiG, he has focused exclusively on VDI and End User Computing (EUC) and oversees strategic technology alliances with leading partners such as Citrix, Microsoft, and Omnissa. Outside of work, Kevin is a devoted family man who enjoys spending time with his wife, two children, and their dog. He enjoys running, cycling and watching sports such as Motorsport & Football/Soccer, especially his son’s team and Leicester City FC.
This post appears through the MedCity Influencers program. Anyone can publish their perspective on business and innovation in healthcare on MedCity News through MedCity Influencers. Click here to find out how.