Healthcare AI tech is notoriously challenging to scale, from the complexity of the applications to the intricacies of the licensing and regulatory environment in an industry where failure can mean life or death.

These unique challenges have meant that the growth of AI in healthcare has been significantly slower than in other industries. Now, driven by the innovative approaches of front-runners, global AI sales in the sector are predicted to exceed $187 billion by 2030. Today, healthcare boasts some of the most effective and transformative approaches to deploying and scaling AI, across a range of use cases and settings.

With this supersonic AI acceleration unfolding now, these pioneers have identified best practices and lessons applicable to every industry.

AI democratizing radiotherapy worldwide

For much of the last decade, the global healthtech innovator Elekta has been developing and commercializing ML-powered systems for radiology and radiation therapy used in the treatment of cancer and brain disorders.  AI and automation are an integral part of “Access 2025,” a company initiative to increase radiotherapy access worldwide especially in underserved markets.

Elekta has partnered to create a dedicated radiotherapy AI center in Amsterdam, the POP-AART lab, that is uncovering ways to use vast amounts of available data to facilitate decision-making at the bedside during treatment. The company is dramatically slashing the time patients spend in treatment — processes and decisions that used to take an hour or even a day can happen in minutes.

Unlocking the power of data globally

With over 4,700 employees spread across 120 countries, Elekta realized that a common scalable data infrastructure was necessary to increase collaboration across teams and ramp up the speed of its AI innovations. To achieve this, the company needed to grab hold of its data pipelines and develop ways to manage data in a secure and distributed fashion. It was also crucial to set up a development and operational environment for machine learning and AI activities that could easily scale.

However, AI has a voracious appetite for data, and Elekta was struggling to get access to large volumes of medical data and medical equipment data required to drive AI development because of privacy concerns. To address this, Elekta established a larger-scale pipeline of anonymized medical data that they could use to drive some of their AI activities.

Moving to the cloud — and tapping a trusted partner

Elekta’s data and research scientists were initially on-prem-centric for data management and compute. The R&D team driving the AI initiative recognized the cloud-based AI infrastructure as a crucial component for effective scaling and uncorking bottlenecks. They found cloud could actually unlock a tremendous number of new opportunities, such as running parallel experiments and multiple scenarios at the same time, scaling GPU capacity and speeding growth and new products and services.

Because AI infrastructure wasn’t their business’s core competence, Elekta discovered that working with trusted partners was the smartest way to build, design and scale a flexible infrastructure to meet and defeat any complexity, from the difficulty of accessing medical data to the broad array of data types and standards, to the complications introduced by proprietary formats that often change over time.

Turning to an end-to-end partner — whether it’s for low-level infrastructure and AI tools, or to offer specific deployable applications — greatly simplified infrastructure and platform integration issues.

LeackStat 2023