TechnologiesAIIndustriesHealthcare & Life Sciences

Tackling the ‘inconvenient truth’ of AI in healthcare

The market is growing fast for Healthcare Artificial Intelligence (AI), estimated to reach over $8 billion by 2026. During this year’s Intelligent Health conference, there was a great deal of talk about the potential of AI in healthcare, but also about its largest challenges: data integration and preparation.

Nature magazine recently called this the ‘inconvenient truth’ of AI in healthcare, noting that data siloes are “severely constraining the ability to provide services to patients across a care continuum within one organization or across organizations … most healthcare organizations lack the data infrastructure required to collect the data needed to optimally train algorithms.”

Our conversations during Intelligent Health reinforce this picture. You can have the greatest AI tools with super-powered algorithms but if you can’t find and utilize the appropriate data then your efforts are doomed to fail. Data preparation and integration are key to future success with AI in healthcare.

Progress is being made, but it’s niche

Intelligent Health showed that there are many exciting and innovative breakthroughs being made. We heard from a range of startups primarily focusing on niche use cases in areas such as radiology image analysis or heart rate and blood circulation.

All of these solutions had one thing in common: by focusing on an isolated problem, they were able to create manageable datasets on which to apply their models and algorithms. Big Pharma, the NHS and other healthcare providers shows that dealing with datasets at scale requires a lot of effort to develop and test applicability of a particular tool for their business processes, research and clinical tasks.

Today, we have a conundrum. We have the innovation and creativity of small players that can bring excellent AI solutions to smaller, well-defined use cases. But to tackle larger challenges, the healthcare industry needs the enterprise strength of larger software vendors that can increase the effectiveness and lower the risks of large scale AI deployments through in-depth data integration and data management capabilities.

Get your data in shape

But we’re beginning to see success stories from Pharma and healthcare providers sharing pieces of their enormous datasets to build or prove its usability. AI is faced with a range of complex Enterprise Information Management challenges and requires an enterprise-strength data management platform – such as OpenText™ Alloy™ – to organize and unlock the value in those massive healthcare datasets.

An enterprise data platform will integrate and transform the data, allowing it to be mapped along the end to end information flow. This ensures the collection of the right data. The data can also be cleansed and transformed, and unstructured content can be enriched with the right industry terminology.

Data management is the heart of AI in Healthcare

Finally, your AI solution can be applied to create new insight delivered at the right time to the right person.

A world of potential

While at Intelligent Health, a senior healthcare executive brought up a very interesting use case. As part of the medical affairs daily job, they must respond to incoming questions from physicians looking for advice on label details. Most of the questions are basic and can be handled by modern call center operations. However, there are still a significant number of questions that need to go directly to Pharma’s medical affairs team. In addition, the call center staff should be able to query the system to have their questions answered. It’s clear that, in this instance, you have to have control of your content and data prior to the application of AI.

The potential for AI to transform the healthcare sector is virtually limitless. Until now, we’ve perhaps spent a little too much time on the algorithms and not enough on the data. We need to flip this situation, and the enterprise information management experience of large software players like OpenText is ideally suited to helping turn this into reality.

If you’d like to know more about the solutions and services that OpenText supply around AI in Healthcare, please visit our website. You can also learn more about our solutions and services for Life Sciences here.

Ferdi Steinmann

Nearly 25 years of experience in driving strategy & commercialization efforts in Biotech & Pharma with an exclusive focus in Life Sciences (LS) strategic planning and industry marketing efforts for enterprise software solutions. Today I am responsible for the LS global industry strategy development at OpenText. I am energized by strategies that deliver on their promises

Related Posts

Back to top button