AI teachings from healthcare: Overcoming difficulty and adopting the cloud

Healthcare AI tech is notoriously problematic to rise, from the complexity of the applications to the complexities of the licensing and regulatory conditions in a sector where failure can mean life or extinction.

These extraordinary challenges have meant that the growth of AI in healthcare has been very slower than in other industries. Now, driven by the creative approaches of front-runners, global AI sales in the sector are anticipated to exceed $187 billion by 2030. Today, healthcare brags some of the most influential and transformative practices for deploying and scaling AI across diverse use cases and environments.

With this high-speed AI acceleration unfolding, these pioneers have recognized best practices and lessons applicable to every enterprise.

AI democratizing radiotherapy globally.

For largely of the last decade, global health tech innovator Elekta has been creating and commercializing ML-powered radiology and radiation therapy techniques to treat cancer and brain conditions. AI and automation are integral to “Access 2025,” a business initiative to increase radiotherapy access worldwide, particularly in underserved areas.

Elekta has partnered to develop a dedicated radiotherapy AI institute in Amsterdam, the POP-AART lab, that is discovering ways to use vast amounts of unrestricted data to facilitate decision-making at the bedside during treatment. The company is dramatically cutting patients’ time in treatment — processes and conclusions that used to take a long time even a day can happen in less time.

Opening the power of data worldwide

With over 4,700 workers spread across 120 countries, Elekta realized that a standard scalable data infrastructure was necessary to improve collaboration across teams and accelerate its AI creations. To achieve this, the company required to grab hold of its data channels and develop ways to manage data securely and distributed. Putting up a development and operational conditions for machine learning and AI activities that could quickly scale was also important.

Nevertheless, AI has a voracious appetite for data. Because of privacy considerations, Elekta was struggling to access large magnitudes of medical data and medical equipment required to drive AI evolution. To address this, Elekta established a larger-scale channel of anonymized medical data that they could employ to guide some of their AI actions.

Shifting to the cloud — and tapping a trusted associate

Elekta’s data and research scientists were originally on-prem-centric for data administration and computing. The R&D team operating the AI initiative recognized the cloud-based AI infrastructure as required for effective scaling and uncorking blockages. They found cloud could unlock many new possibilities, such as running parallel experiments and multiple systems simultaneously, scaling GPU capacity, and speeding growth and new developments and benefits.

Because AI infrastructure wasn’t their company’s core competence, Elekta found that working with trusted partners was the most creative way to build, design and scale a flexible infrastructure to encounter and defeat any complexity, from the hardship of accessing medical data to the broad array of data classes and standards to the complications brought forth by proprietary designs that often change over time.

Shifting to an end-to-end partner — from low-level infrastructure and AI tools or to offering distinctive deployable applications — greatly streamlined infrastructure and platform integration problems.