Discussion around potential applications of artificial intelligence (AI) within healthcare have increased exponentially in the last several years, with excitement over recent advancements in generative AI further fueling these talks. Leaders and decision makers across industries are dipping their toes in the metaphorical waters of AI, investing in new solutions and tools to power the future of their organizations – and healthcare and life science organizations are not exempt. The healthcare and life science industries are particularly primed to see the impact of adopting this technology, yet many find themselves struggling to support complex and ever-evolving AI workflows with their existing data infrastructure.
The adoption, implementation, and integration of AI workflows begins and ends with data. To start, training AI models requires that organizations have large amounts of data they can access easily and process quickly. AI workflows also generate a significant amount of data, which requires quick, safe, and easily accessible data storage management. Many healthcare organizations today have an exponential amount of data at their fingertips, from patient information, to research, to clinical trial data. However, few know how to handle all of this information to fit their organizational needs and those of patients. For organizations to maximize the potential of AI, adequate data storage is a necessary yet often missing piece of the puzzle. Organizations in healthcare and beyond must therefore look toward upgrading their storage capabilities as a first step towards reaping the full benefits of AI and other emerging technologies, as a strong and secure data storage foundation is truly the crux of impactful AI workflows.
What should healthcare organizations look for in data storage?
Performance, flexibility, and reliability are critical to AI functionality and yield more impactful results. Healthcare organizations should prioritize high-performing technology that collects, processes, and stores data efficiently, effectively, and securely, yet allows data to be accessed easily to support future analysis or research. If the current storage infrastructure cannot support these processes, then the potential value of the data cannot be realized. When hospitals and healthcare organizations are able to access, handle, and store data in an effective manner, operations are accelerated, research pipelines are streamlined, and patient outcomes are ultimately improved.
While consistent performance is key, flexibility is just as important. AI tools, techniques, and use cases are rapidly evolving, and so is healthcare. Organizations must invest in infrastructure that can adapt to changes quickly, or else the tools they implement now may be outdated in just a few months or years – just look at how much healthcare technology has advanced in the short time since the pandemic. Choosing technology that can not only support current workflows but can easily adapt to future advancements will set organizations up for long-term success.
Lastly, reliability is essential. When healthcare organizations face a cybersecurity attack, the downtime during which systems cannot access critical patient data directly impacts patient care and is extremely costly. While there’s no way to prevent attacks entirely, having a strong recovery plan that allows organizations to get back up and running as soon as possible after an attack is the best way to minimize interruptions to hospital operations and patient care. With the number of cybersecurity attacks in healthcare growing in quantity and complexity, it’s more important than ever for organizations to have reliable infrastructure so that they can recover quickly should they face a breach. In 2022, healthcare organizations around the world were hit with an average of 1,463 cyberattacks per week – an increase of 74% from 2021 – and this issue is only going to grow worse.
By implementing data storage technology that checks all of these boxes, healthcare organizations will be able to adopt and utilize AI effectively while safeguarding themselves against threats to truly make a difference for researchers, providers, and, most importantly, patients.
How are organizations benefiting from improved data storage?
PAIGE.AI is an organization aiming to revolutionize cancer discovery and treatment through the use of AI in pathology. Most cancer diagnoses depend on pathology, and PAIGE.AI is working to automate this manual, subjective process. However, this requires data storage on an extremely large scale. The organization has petabytes of data from digitizing millions of histological slides, and it uses this data to train AI models. By showing a model thousands of examples of what a specific type of cancer looks like, it learns to recognize that type of cancer when it is shown new images. Once fully trained, AI can assist in accelerating patient diagnosis, reducing or eliminating misdiagnoses, and helping patients get the proper treatments at the right time. Improving its data storage infrastructure has brought PAIGE.AI closer to its goal of transforming the pathology and diagnostics industry and improving the lives of patients with cancer across the country.
In another use case, genetic research was advanced globally after one organization improved their data storage. Chang Gung Memorial Hospital in Taiwan needed better infrastructure to support the AI-driven research at its Center for AI in Medicine. With a massive collection of clinical and genetic data and increasing demand for analysis, the hospital’s existing systems could barely handle AI workloads. After adopting more robust data storage, the hospital was able to increase computing efficiency by 7x and support a greater number of high-performance computing initiatives. This allowed them to conduct multiple projects simultaneously, accelerate research, and produce insights faster that could ultimately improve patient care – all while ensuring that sensitive patient data remained safe and secure.
Nearly every hospital and healthcare organization could benefit from adopting AI, but they need the infrastructure to back it up. PAIGE.AI and Chang Gung Memorial Hospital have already seen major benefits from improving their data storage capabilities. By investing in these solutions to support AI workflows, they’ve been able to improve research pipelines, accelerate processes, and improve patient outcomes. There will be many more advancements in healthcare that have yet to be discovered from maximizing the potential of AI – it’s time we take the right steps to make this a reality.
Jon Kimerle
Jon Kimerle serves as the Epic Alliance Manager at Pure Storage.
